switch from uvicorn to hypercorn

TODO.md

@@ -17,6 +17,8 @@
 23. Do we need cache_limit? We went back and forth on that a lot because we thought it might be causing issues. One problem is it sets it relative to model size. So if you have multiple models loaded in it will take the most recent model size for the cache_limit. This is problematic if you launch DeepSeek -> Llama for example.
 24. further openai/lmstudio api compatibility
 25. Rethink retry logic
+26. Task cancellation. When API http request gets cancelled, it should cancel corresponding task.
+27. Log cleanup - per-module log filters and default to DEBUG log levels
 
 Potential refactors:
 

pyproject.toml

@@ -13,7 +13,6 @@ dependencies = [
     "base58>=2.1.1",
     "cryptography>=45.0.5",
     "fastapi>=0.116.1",
-    "uvicorn>=0.35.0",
     "filelock>=3.18.0",
     "aiosqlite>=0.21.0",
     "networkx>=3.5",
@@ -26,7 +25,6 @@ dependencies = [
     "greenlet>=3.2.4",
     "huggingface-hub>=0.33.4",
     "psutil>=7.0.0",
-    "cobs>=1.2.2",
     "loguru>=0.7.3",
     "textual>=5.3.0",
     "exo_pyo3_bindings", # rust bindings
@@ -35,6 +33,7 @@ dependencies = [
     "mlx>=0.29.3",
     "mlx-lm>=0.28.3",
     "tiktoken>=0.12.0", # required for kimi k2 tokenizer
+    "hypercorn>=0.18.0",
 ]
 
 [project.scripts]

src/exo/main.py

@@ -6,6 +6,7 @@ from typing import Self
 
 import anyio
 from anyio.abc import TaskGroup
+from loguru import logger
 from pydantic import PositiveInt
 
 import exo.routing.topics as topics
@@ -14,7 +15,7 @@ from exo.master.main import Master
 from exo.routing.router import Router, get_node_id_keypair
 from exo.shared.constants import EXO_LOG
 from exo.shared.election import Election, ElectionResult
-from exo.shared.logging import logger, logger_cleanup, logger_setup
+from exo.shared.logging import logger_cleanup, logger_setup
 from exo.shared.types.commands import KillCommand
 from exo.shared.types.common import NodeId, SessionId
 from exo.utils.channels import Receiver, channel

src/exo/master/api.py

@@ -1,21 +1,23 @@
-import asyncio
 import os
 import time
 from collections.abc import AsyncGenerator
-from typing import final
+from typing import cast
 
-import uvicorn
+import anyio
-from anyio import Event as AsyncTaskEvent
 from anyio import create_task_group
 from anyio.abc import TaskGroup
 from fastapi import FastAPI, HTTPException
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import StreamingResponse
 from fastapi.staticfiles import StaticFiles
+from hypercorn.asyncio import serve  # pyright: ignore[reportUnknownVariableType]
+from hypercorn.config import Config
+from hypercorn.typing import ASGIFramework
 from loguru import logger
 
 from exo.shared.apply import apply
 from exo.shared.election import ElectionMessage
+from exo.shared.logging import InterceptLogger
 from exo.shared.models.model_cards import MODEL_CARDS
 from exo.shared.models.model_meta import get_model_meta
 from exo.shared.types.api import (
@@ -46,9 +48,10 @@ from exo.shared.types.state import State
 from exo.shared.types.tasks import ChatCompletionTaskParams
 from exo.shared.types.worker.instances import Instance, InstanceId
 from exo.utils.banner import print_startup_banner
-from exo.utils.channels import Receiver, Sender
+from exo.utils.channels import Receiver, Sender, channel
 from exo.utils.event_buffer import OrderedBuffer
-from exo.worker.engines.mlx.constants import HIDE_THINKING
+
+HIDE_THINKING = False
 
 
 def chunk_to_response(
@@ -76,7 +79,6 @@ async def resolve_model_meta(model_id: str) -> ModelMetadata:
         return await get_model_meta(model_id)
 
 
-@final
 class API:
     def __init__(
         self,
@@ -101,7 +103,7 @@ class API:
         self.port = port
 
         self.paused: bool = False
-        self.paused_ev: AsyncTaskEvent = AsyncTaskEvent()
+        self.paused_ev: anyio.Event = anyio.Event()
 
         self.app = FastAPI()
         self._setup_cors()
@@ -121,7 +123,7 @@ class API:
             name="dashboard",
         )
 
-        self._chat_completion_queues: dict[CommandId, asyncio.Queue[TokenChunk]] = {}
+        self._chat_completion_queues: dict[CommandId, Sender[TokenChunk]] = {}
         self._tg: TaskGroup | None = None
 
     def reset(self, new_session_id: SessionId, result_clock: int):
@@ -135,7 +137,7 @@ class API:
         self.last_completed_election = result_clock
         self.paused = False
         self.paused_ev.set()
-        self.paused_ev = AsyncTaskEvent()
+        self.paused_ev = anyio.Event()
 
     def _setup_cors(self) -> None:
         self.app.add_middleware(
@@ -210,37 +212,40 @@ class API:
     ) -> AsyncGenerator[str, None]:
         """Generate chat completion stream as JSON strings."""
 
-        self._chat_completion_queues[command_id] = asyncio.Queue()
+        try:
+            self._chat_completion_queues[command_id], recv = channel[TokenChunk]()
 
-        finished = False
+            is_thinking = False
-        is_thinking = False
+            with recv as token_chunks:
-        while not finished:
+                async for chunk in token_chunks:
-            # TODO: how long should this timeout be?
+                    if HIDE_THINKING:
-            chunk = await asyncio.wait_for(
+                        if chunk.text == "<think>":
-                self._chat_completion_queues[command_id].get(), timeout=600
+                            is_thinking = True
+                        if chunk.text == "</think>":
+                            is_thinking = False
+                    chunk_response: ChatCompletionResponse = chunk_to_response(
+                        chunk, command_id
+                    )
+                    if not (is_thinking and HIDE_THINKING):
+                        logger.debug(f"chunk_response: {chunk_response}")
+                        yield f"data: {chunk_response.model_dump_json()}\n\n"
+
+                    if chunk.finish_reason is not None:
+                        yield "data: [DONE]\n\n"
+                        break
+
+        except anyio.get_cancelled_exc_class():
+            # TODO: TaskCancelled
+            """
+            self.command_sender.send_nowait(
+                ForwarderCommand(origin=self.node_id, command=command)
             )
-            assert isinstance(chunk, TokenChunk)
+            """
-            # TODO: Do we want this?
+            raise
-            if HIDE_THINKING:
+        finally:
-                if chunk.text == "<think>":
+            command = TaskFinished(finished_command_id=command_id)
-                    chunk.text = "\n"
+            await self._send(command)
-                if chunk.text == "</think>":
+            del self._chat_completion_queues[command_id]
-                    chunk.text = "\n"
-            chunk_response: ChatCompletionResponse = chunk_to_response(
-                chunk, command_id
-            )
-            logger.debug(f"chunk_response: {chunk_response}")
-
-            if not HIDE_THINKING or not is_thinking:
-                yield f"data: {chunk_response.model_dump_json()}\n\n"
-
-            if chunk.finish_reason is not None:
-                yield "data: [DONE]\n\n"
-                finished = True
-
-        command = TaskFinished(finished_command_id=command_id)
-        await self._send(command)
-        del self._chat_completion_queues[command_id]
 
     async def _trigger_notify_user_to_download_model(self, model_id: str) -> None:
         logger.warning(
@@ -298,30 +303,28 @@ class API:
         )
 
     async def run(self):
-        uvicorn_config = uvicorn.Config(
+        cfg = Config()
-            self.app, host="0.0.0.0", port=self.port, access_log=False
+        cfg.bind = f"0.0.0.0:{self.port}"
-        )
+        # nb: shared.logging needs updating if any of this changes
-        uvicorn_server = uvicorn.Server(uvicorn_config)
+        cfg.accesslog = None
+        cfg.errorlog = "-"
+        cfg.logger_class = InterceptLogger
 
         async with create_task_group() as tg:
             self._tg = tg
             logger.info("Starting API")
-            tg.start_soon(uvicorn_server.serve)
             tg.start_soon(self._apply_state)
             tg.start_soon(self._pause_on_new_election)
-            tg.start_soon(self._print_banner_when_ready, uvicorn_server)
+            print_startup_banner(self.port)
+            await serve(
+                cast(ASGIFramework, self.app),
+                cfg,
+                shutdown_trigger=lambda: anyio.sleep_forever(),
+            )
+
         self.command_sender.close()
         self.global_event_receiver.close()
 
-    async def _print_banner_when_ready(self, uvicorn_server: uvicorn.Server):
-        """Wait for the uvicorn server to be ready, then print the startup banner."""
-        # TODO: Is this the best condition to check for?
-        #  The point is this should log when exo is ready.
-        while not uvicorn_server.started:
-            await asyncio.sleep(0.1)
-
-        print_startup_banner(self.port)
-
     async def _apply_state(self):
         with self.global_event_receiver as events:
             async for f_event in events:
@@ -333,7 +336,7 @@ class API:
                         and event.command_id in self._chat_completion_queues
                     ):
                         assert isinstance(event.chunk, TokenChunk)
-                        self._chat_completion_queues[event.command_id].put_nowait(
+                        await self._chat_completion_queues[event.command_id].send(
                             event.chunk
                         )
 

src/exo/shared/constants.py

@@ -1,4 +1,3 @@
-import inspect
 import os
 from pathlib import Path
 
@@ -34,19 +33,3 @@ LIBP2P_COMMANDS_TOPIC = "commands"
 LB_TFLOPS = 2.3
 LB_MEMBW_GBPS = 68
 LB_DISK_GBPS = 1.5
-
-
-# little helper function to get the name of the module that raised the error
-def get_caller_module_name() -> str:
-    frm = inspect.stack()[1]
-    mod = inspect.getmodule(frm[0])
-    if mod is None:
-        return "UNKNOWN MODULE"
-    return mod.__name__
-
-
-def get_error_reporting_message() -> str:
-    return (
-        f"THIS IS A BUG IN THE EXO SOFTWARE, PLEASE REPORT IT AT https://github.com/exo-explore/exo/\n"
-        f"The module that raised the error was: {get_caller_module_name()}"
-    )

src/exo/shared/logging.py

@@ -1,12 +1,39 @@
+import logging
 import sys
 from pathlib import Path
 
+from hypercorn import Config
+from hypercorn.logging import Logger as HypercornLogger
 from loguru import logger
 
 
+class InterceptLogger(HypercornLogger):
+    def __init__(self, config: Config):
+        super().__init__(config)
+        assert self.error_logger
+        # TODO: Decide if we want to provide access logs
+        # assert self.access_logger
+        # self.access_logger.handlers = [_InterceptHandler()]
+        self.error_logger.handlers = [_InterceptHandler()]
+
+
+class _InterceptHandler(logging.Handler):
+    def emit(self, record: logging.LogRecord):
+        try:
+            level = logger.level(record.levelname).name
+        except ValueError:
+            level = record.levelno
+
+        logger.opt(depth=3, exception=record.exc_info).log(level, record.getMessage())
+
+
 def logger_setup(log_file: Path | None, verbosity: int = 0):
     """Set up logging for this process - formatting, file handles, verbosity and output"""
     logger.remove()
+
+    # replace all stdlib loggers with _InterceptHandlers that log to loguru
+    logging.basicConfig(handlers=[_InterceptHandler()], level=0)
+
     if verbosity == 0:
         logger.add(
             sys.__stderr__,  # type: ignore
@@ -37,3 +64,29 @@ def logger_setup(log_file: Path | None, verbosity: int = 0):
 def logger_cleanup():
     """Flush all queues before shutting down so any in-flight logs are written to disk"""
     logger.complete()
+
+
+""" --- TODO: Capture MLX Log output:
+import contextlib
+import sys
+from loguru import logger
+
+class StreamToLogger:
+
+    def __init__(self, level="INFO"):
+        self._level = level
+
+    def write(self, buffer):
+        for line in buffer.rstrip().splitlines():
+            logger.opt(depth=1).log(self._level, line.rstrip())
+
+    def flush(self):
+        pass
+
+logger.remove()
+logger.add(sys.__stdout__)
+
+stream = StreamToLogger()
+with contextlib.redirect_stdout(stream):
+    print("Standard output is sent to added handlers.")
+"""

src/exo/shared/tests/conftest.py

@@ -4,6 +4,8 @@ import asyncio
 from typing import Generator
 
 import pytest
+from _pytest.logging import LogCaptureFixture
+from loguru import logger
 
 from exo.shared.types.memory import Memory
 from exo.shared.types.models import ModelId, ModelMetadata
@@ -41,3 +43,16 @@ def get_pipeline_shard_metadata(
         end_layer=32,
         n_layers=32,
     )
+
+
+@pytest.fixture
+def caplog(caplog: LogCaptureFixture):
+    handler_id = logger.add(
+        caplog.handler,
+        format="{message}",
+        level=0,
+        filter=lambda record: record["level"].no >= caplog.handler.level,
+        enqueue=True,  # Set to 'True' if your test is spawning child processes.
+    )
+    yield caplog
+    logger.remove(handler_id)

src/exo/shared/tests/test_node_id_persistence.py

@@ -1,5 +1,4 @@
 import contextlib
-import logging
 import multiprocessing
 import os
 from multiprocessing import Event, Queue, Semaphore
@@ -8,6 +7,7 @@ from multiprocessing.queues import Queue as QueueT
 from multiprocessing.synchronize import Event as EventT
 from multiprocessing.synchronize import Semaphore as SemaphoreT
 
+from loguru import logger
 from pytest import LogCaptureFixture
 
 from exo.routing.router import get_node_id_keypair
@@ -17,20 +17,13 @@ NUM_CONCURRENT_PROCS = 10
 
 
 def _get_keypair_concurrent_subprocess_task(
-    pid: int, sem: SemaphoreT, ev: EventT, queue: QueueT[bytes]
+    sem: SemaphoreT, ev: EventT, queue: QueueT[bytes]
 ) -> None:
-    try:
+    # synchronise with parent process
-        # synchronise with parent process
+    sem.release()
-        logging.info(msg=f"SUBPROCESS {pid}: Started")
+    # wait to be told to begin simultaneous read
-        sem.release()
+    ev.wait()
-
+    queue.put(get_node_id_keypair().to_protobuf_encoding())
-        # wait to be told to begin simultaneous read
-        ev.wait()
-        logging.info(msg=f"SUBPROCESS {pid}: Reading start")
-        queue.put(get_node_id_keypair().to_protobuf_encoding())
-        logging.info(msg=f"SUBPROCESS {pid}: Reading end")
-    except Exception as e:
-        logging.error(msg=f"SUBPROCESS {pid}: Error encountered: {e}")
 
 
 def _get_keypair_concurrent(num_procs: int) -> bytes:
@@ -41,11 +34,11 @@ def _get_keypair_concurrent(num_procs: int) -> bytes:
     queue: QueueT[bytes] = Queue(maxsize=num_procs)
 
     # make parent process wait for all subprocesses to start
-    logging.info(msg=f"PARENT: Starting {num_procs} subprocesses")
+    logger.info(f"PARENT: Starting {num_procs} subprocesses")
     ps: list[BaseProcess] = []
-    for i in range(num_procs):
+    for _ in range(num_procs):
         p = multiprocessing.get_context("fork").Process(
-            target=_get_keypair_concurrent_subprocess_task, args=(i + 1, sem, ev, queue)
+            target=_get_keypair_concurrent_subprocess_task, args=(sem, ev, queue)
         )
         ps.append(p)
         p.start()
@@ -53,7 +46,7 @@ def _get_keypair_concurrent(num_procs: int) -> bytes:
         sem.acquire()
 
     # start all the sub processes simultaneously
-    logging.info(msg="PARENT: Beginning read")
+    logger.info("PARENT: Beginning read")
     ev.set()
 
     # wait until all subprocesses are done & read results
@@ -62,7 +55,7 @@ def _get_keypair_concurrent(num_procs: int) -> bytes:
 
     # check that the input/output order match, and that
     # all subprocesses end up reading the same file
-    logging.info(msg="PARENT: Checking consistency")
+    logger.info("PARENT: Checking consistency")
     keypair: bytes | None = None
     qsize = 0  # cannot use Queue.qsize due to MacOS incompatibility :(
     while not queue.empty():
@@ -88,7 +81,7 @@ def test_node_id_fetching(caplog: LogCaptureFixture):
     _delete_if_exists(EXO_NODE_ID_KEYPAIR)
     kp = _get_keypair_concurrent(NUM_CONCURRENT_PROCS)
 
-    with caplog.at_level(logging.CRITICAL):  # supress logs
+    with caplog.at_level(101):  # supress logs
         # make sure that continuous fetches return the same value
         for _ in range(reps):
             assert kp == _get_keypair_concurrent(NUM_CONCURRENT_PROCS)

src/exo/utils/banner.py

@@ -1,6 +1,7 @@
 def print_startup_banner(port: int) -> None:
     """Print a prominent startup banner with API endpoint information."""
-    banner = """
+    dashboard_url = f"http://localhost:{port}"
+    banner = f"""
 ╔═══════════════════════════════════════════════════════════════════════╗
 ║                                                                       ║
 ║   ███████╗██╗  ██╗ ██████╗                                            ║
@@ -13,11 +14,7 @@ def print_startup_banner(port: int) -> None:
 ║   Distributed AI Inference Cluster                                    ║
 ║                                                                       ║
 ╚═══════════════════════════════════════════════════════════════════════╝
-"""
 
-    dashboard_url = f"http://localhost:{port}"
-
-    api_info = f"""
 ╔═══════════════════════════════════════════════════════════════════════╗
 ║                                                                       ║
 ║  🌐 Dashboard & API Ready                                             ║
@@ -27,8 +24,7 @@ def print_startup_banner(port: int) -> None:
 ║  Click the URL above to open the dashboard in your browser            ║
 ║                                                                       ║
 ╚═══════════════════════════════════════════════════════════════════════╝
+
 """
 
     print(banner)
-    print(api_info)
-    print()

src/exo/utils/channels.py

@@ -77,9 +77,6 @@ class _MpEndOfStream:
     pass
 
 
-MP_END_OF_STREAM = _MpEndOfStream()
-
-
 class MpState[T]:
     def __init__(self, max_buffer_size: float):
         if max_buffer_size == inf:
@@ -133,7 +130,7 @@ class MpSender[T]:
     def close(self) -> None:
         if not self._state.closed.is_set():
             self._state.closed.set()
-        self._state.buffer.put(MP_END_OF_STREAM)
+        self._state.buffer.put(_MpEndOfStream())
         self._state.buffer.close()
 
     # == unique to Mp channels ==
@@ -177,10 +174,9 @@ class MpReceiver[T]:
 
         try:
             item = self._state.buffer.get(block=False)
-            if item == MP_END_OF_STREAM:
+            if isinstance(item, _MpEndOfStream):
                 self.close()
                 raise EndOfStream
-            assert not isinstance(item, _MpEndOfStream)
             return item
         except Empty:
             raise WouldBlock from None
@@ -193,10 +189,9 @@ class MpReceiver[T]:
             return self.receive_nowait()
         except WouldBlock:
             item = self._state.buffer.get()
-            if item == MP_END_OF_STREAM:
+            if isinstance(item, _MpEndOfStream):
                 self.close()
                 raise EndOfStream from None
-            assert not isinstance(item, _MpEndOfStream)
             return item
 
     # nb: this function will not cancel particularly well

src/exo/worker/download/impl_shard_downloader.py

@@ -20,7 +20,6 @@ def exo_shard_downloader(max_parallel_downloads: int = 8) -> ShardDownloader:
 
 async def build_base_shard(model_id: str) -> ShardMetadata:
     model_meta = await get_model_meta(model_id)
-    # print(f"build_base_shard {model_id=} {model_meta=}")
     return PipelineShardMetadata(
         model_meta=model_meta,
         device_rank=0,
@@ -92,10 +91,8 @@ class CachedShardDownloader(ShardDownloader):
         self, shard: ShardMetadata, config_only: bool = False
     ) -> Path:
         if (shard.model_meta.model_id, shard) in self.cache:
-            # print(f"ensure_shard cache hit {shard=}")
             return self.cache[(shard.model_meta.model_id, shard)]
 
-        # print(f"ensure_shard cache miss {shard=}")
         target_dir = await self.shard_downloader.ensure_shard(shard, config_only)
         self.cache[(shard.model_meta.model_id, shard)] = target_dir
         return target_dir
@@ -135,7 +132,6 @@ class ResumableShardDownloader(ShardDownloader):
     ) -> Path:
         allow_patterns = ["config.json"] if config_only else None
 
-        # print(f"ensure_shard {shard=} {config_only=} {allow_patterns=}")
         target_dir, _ = await download_shard(
             shard,
             self.on_progress_wrapper,
@@ -147,7 +143,6 @@ class ResumableShardDownloader(ShardDownloader):
     async def get_shard_download_status(
         self,
     ) -> AsyncIterator[tuple[Path, RepoDownloadProgress]]:
-        # print("get_shard_download_status")
         async def _status_for_model(
             model_id: str,
         ) -> tuple[Path, RepoDownloadProgress]:
@@ -165,9 +160,8 @@ class ResumableShardDownloader(ShardDownloader):
 
         for task in asyncio.as_completed(tasks):
             try:
-                result = await task
+                yield await task
-                path, progress = result
+            # TODO: except Exception
-                yield (path, progress)
             except Exception as e:
                 print("Error downloading shard:", e)
 

src/exo/worker/engines/mlx/constants.py

@@ -14,5 +14,3 @@ TEMPERATURE: float = 1.0
 
 # TODO: We should really make this opt-in, but Kimi requires trust_remote_code=True
 TRUST_REMOTE_CODE: bool = True
-# TODO: Do we really want this?
-HIDE_THINKING: bool = False

uv.lock

@@ -258,21 +258,6 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/0a/4c/925909008ed5a988ccbb72dcc897407e5d6d3bd72410d69e051fc0c14647/charset_normalizer-3.4.4-py3-none-any.whl", hash = "sha256:7a32c560861a02ff789ad905a2fe94e3f840803362c84fecf1851cb4cf3dc37f", size = 53402, upload-time = "2025-10-14T04:42:31.76Z" },
 ]
 
-[[package]]
-name = "click"
-version = "8.3.0"
-source = { registry = "https://pypi.org/simple" }
-sdist = { url = "https://files.pythonhosted.org/packages/46/61/de6cd827efad202d7057d93e0fed9294b96952e188f7384832791c7b2254/click-8.3.0.tar.gz", hash = "sha256:e7b8232224eba16f4ebe410c25ced9f7875cb5f3263ffc93cc3e8da705e229c4", size = 276943, upload-time = "2025-09-18T17:32:23.696Z" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/db/d3/9dcc0f5797f070ec8edf30fbadfb200e71d9db6b84d211e3b2085a7589a0/click-8.3.0-py3-none-any.whl", hash = "sha256:9b9f285302c6e3064f4330c05f05b81945b2a39544279343e6e7c5f27a9baddc", size = 107295, upload-time = "2025-09-18T17:32:22.42Z" },
-]
-
-[[package]]
-name = "cobs"
-version = "1.2.2"
-source = { registry = "https://pypi.org/simple" }
-sdist = { url = "https://files.pythonhosted.org/packages/34/ef/ea149311227a4fc3160cc885fce06da7c7d76782a308ef070b8065c69953/cobs-1.2.2.tar.gz", hash = "sha256:dbdd5e32111d72786f83d0c269215dcd6ac629b1ac1962c6878221f3b2ca98da", size = 14582, upload-time = "2025-07-20T01:08:35.434Z" }
-
 [[package]]
 name = "cryptography"
 version = "46.0.3"
@@ -331,13 +316,13 @@ dependencies = [
     { name = "anyio", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "base58", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "bidict", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
-    { name = "cobs", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "cryptography", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "exo-pyo3-bindings", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "fastapi", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "filelock", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "greenlet", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "huggingface-hub", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+    { name = "hypercorn", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "loguru", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "mlx", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "mlx-lm", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
@@ -354,7 +339,6 @@ dependencies = [
     { name = "tiktoken", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "typeguard", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "types-aiofiles", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
-    { name = "uvicorn", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
 ]
 
 [package.dev-dependencies]
@@ -373,13 +357,13 @@ requires-dist = [
     { name = "anyio", specifier = "==4.11.0" },
     { name = "base58", specifier = ">=2.1.1" },
     { name = "bidict", specifier = ">=0.23.1" },
-    { name = "cobs", specifier = ">=1.2.2" },
     { name = "cryptography", specifier = ">=45.0.5" },
     { name = "exo-pyo3-bindings", editable = "rust/exo_pyo3_bindings" },
     { name = "fastapi", specifier = ">=0.116.1" },
     { name = "filelock", specifier = ">=3.18.0" },
     { name = "greenlet", specifier = ">=3.2.4" },
     { name = "huggingface-hub", specifier = ">=0.33.4" },
+    { name = "hypercorn", specifier = ">=0.18.0" },
     { name = "loguru", specifier = ">=0.7.3" },
     { name = "mlx", specifier = ">=0.29.3" },
     { name = "mlx-lm", specifier = ">=0.28.3" },
@@ -396,7 +380,6 @@ requires-dist = [
     { name = "tiktoken", specifier = ">=0.12.0" },
     { name = "typeguard", specifier = ">=4.4.4" },
     { name = "types-aiofiles", specifier = ">=24.1.0.20250708" },
-    { name = "uvicorn", specifier = ">=0.35.0" },
 ]
 
 [package.metadata.requires-dev]
@@ -557,6 +540,19 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/04/4b/29cac41a4d98d144bf5f6d33995617b185d14b22401f75ca86f384e87ff1/h11-0.16.0-py3-none-any.whl", hash = "sha256:63cf8bbe7522de3bf65932fda1d9c2772064ffb3dae62d55932da54b31cb6c86", size = 37515, upload-time = "2025-04-24T03:35:24.344Z" },
 ]
 
+[[package]]
+name = "h2"
+version = "4.3.0"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "hpack", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+    { name = "hyperframe", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/1d/17/afa56379f94ad0fe8defd37d6eb3f89a25404ffc71d4d848893d270325fc/h2-4.3.0.tar.gz", hash = "sha256:6c59efe4323fa18b47a632221a1888bd7fde6249819beda254aeca909f221bf1", size = 2152026, upload-time = "2025-08-23T18:12:19.778Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/69/b2/119f6e6dcbd96f9069ce9a2665e0146588dc9f88f29549711853645e736a/h2-4.3.0-py3-none-any.whl", hash = "sha256:c438f029a25f7945c69e0ccf0fb951dc3f73a5f6412981daee861431b70e2bdd", size = 61779, upload-time = "2025-08-23T18:12:17.779Z" },
+]
+
 [[package]]
 name = "hf-xet"
 version = "1.2.0"
@@ -583,6 +579,15 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/92/68/89ac4e5b12a9ff6286a12174c8538a5930e2ed662091dd2572bbe0a18c8a/hf_xet-1.2.0-cp37-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:a55558084c16b09b5ed32ab9ed38421e2d87cf3f1f89815764d1177081b99865", size = 3508920, upload-time = "2025-10-24T19:04:26.927Z" },
 ]
 
+[[package]]
+name = "hpack"
+version = "4.1.0"
+source = { registry = "https://pypi.org/simple" }
+sdist = { url = "https://files.pythonhosted.org/packages/2c/48/71de9ed269fdae9c8057e5a4c0aa7402e8bb16f2c6e90b3aa53327b113f8/hpack-4.1.0.tar.gz", hash = "sha256:ec5eca154f7056aa06f196a557655c5b009b382873ac8d1e66e79e87535f1dca", size = 51276, upload-time = "2025-01-22T21:44:58.347Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/07/c6/80c95b1b2b94682a72cbdbfb85b81ae2daffa4291fbfa1b1464502ede10d/hpack-4.1.0-py3-none-any.whl", hash = "sha256:157ac792668d995c657d93111f46b4535ed114f0c9c8d672271bbec7eae1b496", size = 34357, upload-time = "2025-01-22T21:44:56.92Z" },
+]
+
 [[package]]
 name = "huggingface-hub"
 version = "0.36.0"
@@ -602,6 +607,30 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/cb/bd/1a875e0d592d447cbc02805fd3fe0f497714d6a2583f59d14fa9ebad96eb/huggingface_hub-0.36.0-py3-none-any.whl", hash = "sha256:7bcc9ad17d5b3f07b57c78e79d527102d08313caa278a641993acddcb894548d", size = 566094, upload-time = "2025-10-23T12:11:59.557Z" },
 ]
 
+[[package]]
+name = "hypercorn"
+version = "0.18.0"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "h11", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+    { name = "h2", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+    { name = "priority", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+    { name = "wsproto", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/44/01/39f41a014b83dd5c795217362f2ca9071cf243e6a75bdcd6cd5b944658cc/hypercorn-0.18.0.tar.gz", hash = "sha256:d63267548939c46b0247dc8e5b45a9947590e35e64ee73a23c074aa3cf88e9da", size = 68420, upload-time = "2025-11-08T13:54:04.78Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/93/35/850277d1b17b206bd10874c8a9a3f52e059452fb49bb0d22cbb908f6038b/hypercorn-0.18.0-py3-none-any.whl", hash = "sha256:225e268f2c1c2f28f6d8f6db8f40cb8c992963610c5725e13ccfcddccb24b1cd", size = 61640, upload-time = "2025-11-08T13:54:03.202Z" },
+]
+
+[[package]]
+name = "hyperframe"
+version = "6.1.0"
+source = { registry = "https://pypi.org/simple" }
+sdist = { url = "https://files.pythonhosted.org/packages/02/e7/94f8232d4a74cc99514c13a9f995811485a6903d48e5d952771ef6322e30/hyperframe-6.1.0.tar.gz", hash = "sha256:f630908a00854a7adeabd6382b43923a4c4cd4b821fcb527e6ab9e15382a3b08", size = 26566, upload-time = "2025-01-22T21:41:49.302Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/48/30/47d0bf6072f7252e6521f3447ccfa40b421b6824517f82854703d0f5a98b/hyperframe-6.1.0-py3-none-any.whl", hash = "sha256:b03380493a519fce58ea5af42e4a42317bf9bd425596f7a0835ffce80f1a42e5", size = 13007, upload-time = "2025-01-22T21:41:47.295Z" },
+]
+
 [[package]]
 name = "idna"
 version = "3.11"
@@ -926,6 +955,15 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/54/20/4d324d65cc6d9205fabedc306948156824eb9f0ee1633355a8f7ec5c66bf/pluggy-1.6.0-py3-none-any.whl", hash = "sha256:e920276dd6813095e9377c0bc5566d94c932c33b27a3e3945d8389c374dd4746", size = 20538, upload-time = "2025-05-15T12:30:06.134Z" },
 ]
 
+[[package]]
+name = "priority"
+version = "2.0.0"
+source = { registry = "https://pypi.org/simple" }
+sdist = { url = "https://files.pythonhosted.org/packages/f5/3c/eb7c35f4dcede96fca1842dac5f4f5d15511aa4b52f3a961219e68ae9204/priority-2.0.0.tar.gz", hash = "sha256:c965d54f1b8d0d0b19479db3924c7c36cf672dbf2aec92d43fbdaf4492ba18c0", size = 24792, upload-time = "2021-06-27T10:15:05.487Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/5e/5f/82c8074f7e84978129347c2c6ec8b6c59f3584ff1a20bc3c940a3e061790/priority-2.0.0-py3-none-any.whl", hash = "sha256:6f8eefce5f3ad59baf2c080a664037bb4725cd0a790d53d59ab4059288faf6aa", size = 8946, upload-time = "2021-06-27T10:15:03.856Z" },
+]
+
 [[package]]
 name = "propcache"
 version = "0.4.1"
@@ -1524,16 +1562,15 @@ wheels = [
 ]
 
 [[package]]
-name = "uvicorn"
+name = "wsproto"
-version = "0.38.0"
+version = "1.3.2"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
-    { name = "click", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "h11", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/cb/ce/f06b84e2697fef4688ca63bdb2fdf113ca0a3be33f94488f2cadb690b0cf/uvicorn-0.38.0.tar.gz", hash = "sha256:fd97093bdd120a2609fc0d3afe931d4d4ad688b6e75f0f929fde1bc36fe0e91d", size = 80605, upload-time = "2025-10-18T13:46:44.63Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/c7/79/12135bdf8b9c9367b8701c2c19a14c913c120b882d50b014ca0d38083c2c/wsproto-1.3.2.tar.gz", hash = "sha256:b86885dcf294e15204919950f666e06ffc6c7c114ca900b060d6e16293528294", size = 50116, upload-time = "2025-11-20T18:18:01.871Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/ee/d9/d88e73ca598f4f6ff671fb5fde8a32925c2e08a637303a1d12883c7305fa/uvicorn-0.38.0-py3-none-any.whl", hash = "sha256:48c0afd214ceb59340075b4a052ea1ee91c16fbc2a9b1469cca0e54566977b02", size = 68109, upload-time = "2025-10-18T13:46:42.958Z" },
+    { url = "https://files.pythonhosted.org/packages/a4/f5/10b68b7b1544245097b2a1b8238f66f2fc6dcaeb24ba5d917f52bd2eed4f/wsproto-1.3.2-py3-none-any.whl", hash = "sha256:61eea322cdf56e8cc904bd3ad7573359a242ba65688716b0710a5eb12beab584", size = 24405, upload-time = "2025-11-20T18:18:00.454Z" },
 ]
 
 [[package]]