# exo changelog

> exo release notes.

- Vendor: exo
- Category: AI
- Official site: https://github.com/exo-explore/exo
- Tracked by: What's New (https://whatsnew.fyi/product/exo)
- Harvested from: GitHub (exo-explore/exo)
- Entries below: 10 (newest first)

What's New is an index, not a publisher: every entry below links to the vendor's own release notes, which are the authoritative source. Entries are labelled where they are hand-curated sample data, pre-releases, or drawn from a secondary source such as a developer blog.

## Releases

### v1.0.71

- Date: 2026-04-23
- Version: v1.0.71
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.71
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.71

#### EXO v1.0.71 Release Notes

This is a small patch release, with better defaults for sampling and bug fixes for M5 series Macs and RDMA, along with support for the new Kimi K2.6.

##### Models
- Add Kimi K2.6, including support for multimodality ([#1952](https://github.com/exo-explore/exo/pull/1952))
- Add model cards for Qwen3.6 27B, GLM 5.1 and Kimi K2.6 quantisations ([#1970](https://github.com/exo-explore/exo/pull/1970))


##### Performance
- Add recommended sampling defaults (from model providers) to improve model output quality, including min_p and top_k ([#1947](https://github.com/exo-explore/exo/pull/1947))
- Prevent JACCL all_sum from corrupting output ([#1952](https://github.com/exo-explore/exo/pull/1952))

##### Quality of Life
- Add a Pi integration tab! ([#1925](https://github.com/exo-explore/exo/pull/1925))
- Handle multiple tool calls correctly in the Claude and Responses APIs ([#1952](https://github.com/exo-explore/exo/pull/1952))
- Add a friendly context message for bug reporting ([#1959](https://github.com/exo-explore/exo/pull/1959))

##### Bug Fixes
- Fix vision models on M5 series Macs by fixing addmm on NAX ([#1927](https://github.com/exo-explore/exo/pull/1927))
- Fix prefix cache for Qwen3.6 ([#1952](https://github.com/exo-explore/exo/pull/1952))
- Prevent complete downloads from losing their Completed status ([#1956](https://github.com/exo-explore/exo/pull/1956), [#1958](https://github.com/exo-explore/exo/pull/1958))
- Populate the image cache from events to prevent crashes ([#1961](https://github.com/exo-explore/exo/pull/1961))
- Fix DeepSeekV32 parsing and tensor parallelism for Gemma 4 models with effective parameters ([#1967](https://github.com/exo-explore/exo/pull/1967))

##### New Contributors
And a thank you to [@nhwaani](https://github.com/nhwaani) who made their first contribution here!

**Full Changelog**: [https://github.com/exo-explore/exo/compare/v1.0.70...v1.0.71](https://github.com/exo-explore/exo/compare/v1.0.70...v1.0.71)

### v1.0.70

- Date: 2026-04-17
- Version: v1.0.70
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.70
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.70

#### EXO v1.0.70 Release Notes

This release ships with multimodality and major enhancements for memory usage in long context use cases (e.g. OpenClaw and OpenCode), as well as updated model support and QOL features.

##### Models
- Add support for Gemma 4 ([#1851](https://github.com/exo-explore/exo/pull/1851), [#1891](https://github.com/exo-explore/exo/pull/1891))
- Add support for Minimax M2.7 ([#1884](https://github.com/exo-explore/exo/pull/1884))
- Add support for Qwen3.6 ([#1907](https://github.com/exo-explore/exo/pull/1907), [#1917](https://github.com/exo-explore/exo/pull/1917))

##### Performance
- Implement Flash Attention for Qwen3.5 and Gemma 4, reducing peak memory consumption by 3-6x ([#1886](https://github.com/exo-explore/exo/pull/1886))
- Fix memory leaks in Rotating and Arrays cache ([#1819](https://github.com/exo-explore/exo/pull/1819), [#1835](https://github.com/exo-explore/exo/pull/1835))
- Improve KV prefix cache hit rates ([#1817](https://github.com/exo-explore/exo/pull/1817))
- Garbage collect KV cache on eviction ([#1832](https://github.com/exo-explore/exo/pull/1832))

##### Multimodality
- Add support for vision models, making use of vision processors in Qwen3.5, Kimi K2.5 and Gemma 4 ([#1802](https://github.com/exo-explore/exo/pull/1802), [#1893](https://github.com/exo-explore/exo/pull/1893))
- Handle PDFs by sending both text and image ([#1822](https://github.com/exo-explore/exo/pull/1822), [#1865](https://github.com/exo-explore/exo/pull/1865))

##### API
- Add integration helpers to display one-click setups for connecting to tools such as OpenCode, n8n and OpenClaw ([#1810](https://github.com/exo-explore/exo/pull/1810))
- Add Firefox AI sidebar (?q=) support to dashboard ([#1814](https://github.com/exo-explore/exo/pull/1814))
- Expose /state/paths for easier lookup of pieces of state ([#1796](https://github.com/exo-explore/exo/pull/1796))
- Improve stats and usage reporting ([#1836](https://github.com/exo-explore/exo/pull/1836), [#1848](https://github.com/exo-explore/exo/pull/1848), [#1885](https://github.com/exo-explore/exo/pull/1885), [#1899](https://github.com/exo-explore/exo/pull/1899))
- Prevent timing out on clients during long prefills ([#1803](https://github.com/exo-explore/exo/pull/1803), [#1828](https://github.com/exo-explore/exo/pull/1828))
- Allow copying even on http, as a QOL for Tailscale users and other non localhost serving ([#1902](https://github.com/exo-explore/exo/pull/1902))

##### Quality of Life
- Allow pausing and deleting active downloads ([#1829](https://github.com/exo-explore/exo/pull/1829))
- Add the ability to edit environment variables in MacOS app ([#1852](https://github.com/exo-explore/exo/pull/1852), [#1869](https://github.com/exo-explore/exo/pull/1869), [#1901](https://github.com/exo-explore/exo/pull/1901))
- Prevent failed instance retry loops ([#1763](https://github.com/exo-explore/exo/pull/1763))
- Truncate overly verbose bits of text in logs ([#1854](https://github.com/exo-explore/exo/pull/1854))

##### Bug Fixes
- Clean up processes nicely to prevent zombie processes and holding RDMA resources ([#1889](https://github.com/exo-explore/exo/pull/1889))
- Fix out of order event crashes on startup ([#1894](https://github.com/exo-explore/exo/pull/1894))
- Drain tokens silently skipped in thinking parsing, prevent malformed tool calls in certain models ([#1898](https://github.com/exo-explore/exo/pull/1898))
- Gracefully catch errors in downloads ([#1846](https://github.com/exo-explore/exo/pull/1846))
- Prevent several fatal crashes ([#1856](https://github.com/exo-explore/exo/pull/1856), [#1827](https://github.com/exo-explore/exo/pull/1827))
- Use configured api_port for IP connectivity probes ([#1877](https://github.com/exo-explore/exo/pull/1877))

##### New Contributors
We had some more new faces in this release. Thank you for your contributions!
* [@kaiisfree](https://github.com/kaiisfree)
* [@MikkoParkkola](https://github.co

_[Truncated at 4000 characters — full notes: https://github.com/exo-explore/exo/releases/tag/v1.0.70]_

### v1.0.69

- Date: 2026-03-27
- Version: v1.0.69
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.69
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.69

#### EXO v1.0.69 Release Notes

This release ships with continuous batching, Qwen3.5 support and support for M5 Pro/Max chips, as well as a host of quality of life improvements and bug fixes.

**Continuous batching** is on by default, enabling you to run multiple requests in parallel for significantly higher throughput. EXO will automatically batch together inference requests, on single node and multi-node instances including RDMA instances. This is particularly useful for agentic workflows where multiple agents can run in parallel.

##### Models

- Add support for Qwen3.5 ([#1644](https://github.com/exo-explore/exo/pull/1644))
- Add support for Nemotron sharding ([#1693](https://github.com/exo-explore/exo/pull/1693))
- Add default model cards for DeepSeek v3.2 ([#1769](https://github.com/exo-explore/exo/pull/1769))

##### API

- Add POST /v1/cancel/{command_id} endpoint for cancelling ongoing text generations ([#1579](https://github.com/exo-explore/exo/pull/1579))
- Add reasoning params to chat completions and responses APIs ([#1654](https://github.com/exo-explore/exo/pull/1654))
- Add `repetition_penalty` and `repetition_context_size` to chat completions ([#1665](https://github.com/exo-explore/exo/pull/1665))

##### Performance

- Continuous batching ([#1642](https://github.com/exo-explore/exo/pull/1579), [#1632](https://github.com/exo-explore/exo/pull/1632), [#1777](https://github.com/exo-explore/exo/pull/1777))
- Better pipeline parallel prefill that splits the prompt into chunks and overlaps computation and communication. This makes pipeline parallel prefill up to 1.98x faster on 2 nodes ([#1587](https://github.com/exo-explore/exo/pull/1587), [#1629](https://github.com/exo-explore/exo/pull/1629))

##### Quality of Life

- Support trace deletion in dashboard ([#1628](https://github.com/exo-explore/exo/pull/1628))
- Make mini topology sidebar navigate to home on click ([#1616](https://github.com/exo-explore/exo/pull/1616))
- Show feedback that the model was successfully added when adding custom models from Huggingface ([#1661](https://github.com/exo-explore/exo/pull/1661))
- Enable global model search from huggingface, not just `mlx-community` ([#1661](https://github.com/exo-explore/exo/pull/1661))
- Mobile-friendly UI ([#1677](https://github.com/exo-explore/exo/pull/1677))
- Include power usage in exo-bench responses, enabling benchmarks to capture energy usage ([#1693](https://github.com/exo-explore/exo/pull/1693))
- Prefer nodes with more of the model downloaded when placing an instance ([#1767](https://github.com/exo-explore/exo/pull/1767), [#1795](https://github.com/exo-explore/exo/pull/1795))
- Sync custom model cards across nodes ([#1768](https://github.com/exo-explore/exo/pull/1768))
- Add `--bootstrap-peers` and `--libp2p-port` for static peer discovery, bypassing mDNS which Is useful in environments where mDNS is unavailable ([#1690](https://github.com/exo-explore/exo/pull/1690))

##### Image Generation (Experimental)

- Update mflux to 0.16.9 ([#1751](https://github.com/exo-explore/exo/pull/1751))
- Support image generation cancellation ([#1774](https://github.com/exo-explore/exo/pull/1774))

##### Bug Fixes

- Upgrade macmon to fix macmon errors on M5 Pro/Max. This fixes an issue where M5 Pro/Max would not report memory or GPU usage stats and therefore could not participate in clusters ([#1747](https://github.com/exo-explore/exo/pull/1747), [#1797](https://github.com/exo-explore/exo/pull/1797))
- Use tmpdir for MLX distributed coordination file, preventing local network access permission issues ([#1624](https://github.com/exo-explore/exo/pull/1624))
- Fix BrokenResourceError crash when immediately loading a model on start ([#1637](https://github.com/exo-explore/exo/pull/1637))
- Emit error chunks when a runner crashes in the middle of a request, preventing streams hanging forever when runners crash ([#1645](https://github.com/exo-explore/exo/pull/1645))
- Fix copy co

_[Truncated at 4000 characters — full notes: https://github.com/exo-explore/exo/releases/tag/v1.0.69]_

### v1.0.68

- Date: 2026-02-25
- Version: v1.0.68
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.68
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.68

#### EXO v1.0.68 Release Notes

This is the biggest EXO release to date. We wanted to make sure we address the stability issues users were running into on previous versions and we think we've achieved that with this release. This release also comes with a whole load of new features and UX improvements, full list below. Thank you to everyone who submitted bug reports over the past few weeks - it helps us to improve EXO much faster.

##### Models

- Add support for custom models from Huggingface ([#1368](https://github.com/exo-explore/exo/pull/1368))
- Add support for Qwen3-Coder-Next ([#1367](https://github.com/exo-explore/exo/pull/1367))
- Add support for Step 3.5 Flash ([#1460](https://github.com/exo-explore/exo/pull/1460))
- Add support for GLM 5 ([#1526](https://github.com/exo-explore/exo/pull/1526)), ([#1529](https://github.com/exo-explore/exo/pull/1529))
- Add support for MiniMax M2.5 ([#1514](https://github.com/exo-explore/exo/pull/1514))

##### API

- Add support for Claude Messages API, enabling tools like Claude Code ([#1167](https://github.com/exo-explore/exo/pull/1167))
- Add support for OpenAI Responses API ([#1167](https://github.com/exo-explore/exo/pull/1167))
- Add usage and generation stats to API, enabling clients like OpenCode to consume stats including prompt tokens, completion tokens and total tokens ([#1333](https://github.com/exo-explore/exo/pull/1333)), ([#1461](https://github.com/exo-explore/exo/pull/1461))
- Cancel text generation when API request is closed ([#1276](https://github.com/exo-explore/exo/pull/1276))
- Add support for Ollama API ([#1560](https://github.com/exo-explore/exo/pull/1560))

##### Web Dashboard

- Add redesigned model picker modal ([#1369](https://github.com/exo-explore/exo/pull/1369)), ([#1377](https://github.com/exo-explore/exo/pull/1377)), ([#1440](https://github.com/exo-explore/exo/pull/1440)), ([#1470](https://github.com/exo-explore/exo/pull/1470))
- Display alternative tokens / logprobs visualizer in chat responses ([#1180](https://github.com/exo-explore/exo/pull/1180))
- Redesign downloads page as model x node table ([#1465](https://github.com/exo-explore/exo/pull/1465)), ([#1589](https://github.com/exo-explore/exo/pull/1589)), ([#1581](https://github.com/exo-explore/exo/pull/1581))
- Add prefill progress bar for long prompts ([#1181](https://github.com/exo-explore/exo/pull/1181)), ([#1557](https://github.com/exo-explore/exo/pull/1557))
- A new onboarding flow when running EXO for the first time ([#1533](https://github.com/exo-explore/exo/pull/1533))
- Automatic model selection / model recommendations in web dashboard ([#1590](https://github.com/exo-explore/exo/pull/1590))

##### Quality of Life

- Show a more informative message in macOS app when installing network location ([#1309](https://github.com/exo-explore/exo/pull/1309))
- Migrate model cards to .toml files ([#1354](https://github.com/exo-explore/exo/pull/1354))
- Clean up exo gracefully on shutdown, preventing memory not being cleaned up on exit ([#1388](https://github.com/exo-explore/exo/pull/1388))
- Make topology updates more responsive by yielding from reachability checks instead of waiting for all checks ([#1427](https://github.com/exo-explore/exo/pull/1427))
- Allow typing in chat input while response is generating ([#1433](https://github.com/exo-explore/exo/pull/1433))
- Add log rotation, now exo logs get written to `~/.exo/exo_logs` ([#1438](https://github.com/exo-explore/exo/pull/1438)), ([#1439](https://github.com/exo-explore/exo/pull/1439)), ([#1442](https://github.com/exo-explore/exo/pull/1442))
- Distinguish between model fits in available memory and fits in total memory in model picker ([#1441](https://github.com/exo-explore/exo/pull/1441) h/t [@Hmbown](https://github.com/Hmbown))
- Add `enable_thinking` toggle for models that support thinking/non-thinking ([#1457](https://github.com/exo-explore/exo/pull/1457))
- Show a warning in the web dashboard when macOS ve

_[Truncated at 4000 characters — full notes: https://github.com/exo-explore/exo/releases/tag/v1.0.68]_

### v1.0.67

- Date: 2026-01-28
- Version: v1.0.67
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.67
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.67

#### EXO v1.0.67 Release Notes

This release adds support for Kimi-K2.5 and tensor sharding support for MiniMax M2.1.

##### Models

- Add support for Kimi K2.5 ([#1302](https://github.com/exo-explore/exo/pull/1302))
- Add tensor sharding support for MiniMax M2.1 ([#1299](https://github.com/exo-explore/exo/pull/1299))

**Full Changelog:** https://github.com/exo-explore/exo/compare/v1.0.66...v1.0.67

### v1.0.66

- Date: 2026-01-26
- Version: v1.0.66
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.66
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.66

#### EXO v1.0.66 Release Notes

This is a stability release that fixes a regression with RDMA / Tensor Parallelism where models were getting stuck in `LOADING` state. It also fixes a download edge case with GLM 4.7 Flash and nodes getting stuck in `UNKNOWN` state / zombie states after periods of inactivity.

All models have been confirmed working with RDMA / Tensor Parallel on various configurations (including Mac Minis, MacBooks and Mac Studios). Thank you to users who reported bugs to help us resolve these issues - it helps a lot!

##### Bug Fixes

- Use EXO shard instead of upstream shard for all models, loading models layer-by-layer, fixing models getting stuck in `LOADING` state e.g.`GLM-4.7-Flash-4bit`, `gpt-oss-120b-MXFP4-Q8` and `Qwen3-Coder-480B-A35B-Instruct-8bit`. Also fixes memory not being released when an instance is deleted. ([#1291](https://github.com/exo-explore/exo/pull/1291))
- Fix downloads getting stuck when model files change in Huggingface repo e.g. `GLM-4.7-Flash-4bit` which was updated upstream on Jan 25 ([#1290](https://github.com/exo-explore/exo/pull/1290))
- Always publish info gatherer events, preventing nodes getting stuck in `UNKNOWN` state / zombie states after a period of inactivity ([#1283](https://github.com/exo-explore/exo/pull/1283))
- Fix tool calls with empty text content ([#1292](https://github.com/exo-explore/exo/pull/1292))

##### MLX

- Upgrade `mlx-lm` to `0.30.5`

**Full Changelog:** https://github.com/exo-explore/exo/compare/v1.0.65...v1.0.66

### v1.0.65

- Date: 2026-01-24
- Version: v1.0.65
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.65
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.65

#### EXO v1.0.65 Release Notes

This release ships with stability fixes for RDMA and long running clusters (fixing the _FAILED_ -> _PREPARING_ loops for some clusters), as well as QoL features for managing downloads and a fix for GPT-OSS tool calling.

##### UX

- Add download and delete buttons to downloads UI ([#1236](https://github.com/exo-explore/exo/pull/1236), ([#1237](https://github.com/exo-explore/exo/pull/1237)))

##### Bug Fixes

- Fix parsing logic for GPT-OSS tool calling ([#1271](https://github.com/exo-explore/exo/pull/1271))
- Fix Thunderbolt bridge cycle detection to include 2-node cycles ([#1261](https://github.com/exo-explore/exo/pull/1261))
- Fix placement filter to use subset matching instead of exact match ([#1265](https://github.com/exo-explore/exo/pull/1265))
- Fix instance port assignment, improving stability for RDMA clusters ([#1268](https://github.com/exo-explore/exo/pull/1268))
- Deprioritise uncertain ethernet devices, improving stability for RDMA clusters ([#1267](https://github.com/exo-explore/exo/pull/1267))
- Restore Thunderbolt Bridge LaunchDaemon, since TB bridge gets enabled by macOS on reboot automatically ([#1270](https://github.com/exo-explore/exo/pull/1270))
- Add back the EXO network profile creation to the LaunchDaemon ([#1277](https://github.com/exo-explore/exo/pull/1277))

**Full Changelog:** https://github.com/exo-explore/exo/compare/v1.0.64...v1.0.65

### v1.0.64

- Date: 2026-01-23
- Version: v1.0.64
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.64
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.64

#### EXO v1.0.64 Release Notes

This release comes with [support for GLM-4.7-Flash](https://x.com/exolabs/status/2013458583023698014), IP-less RDMA discovery (removing the need for custom network locations) and OpenAI-compatible tool calling via the API. It also includes bug fixes for auto parallelism, fixing various models including Qwen, GPT-OSS and MiniMax that were getting stuck in LOADING / WARMING UP, as well as better error messages when things go wrong.

##### Model Support

- Added support for GLM-4.7-Flash ([#1214](https://github.com/exo-explore/exo/pull/1214))

##### API

- Added tool calling support to the OpenAI-compatible chat completions API ([#1233](https://github.com/exo-explore/exo/pull/1233))

##### UX

- Add proxy and custom SSL certificate support for corporate networks ([#1189](https://github.com/exo-explore/exo/pull/1189))
- More responsive node info by splitting information sources and state ([#928](https://github.com/exo-explore/exo/pull/928), [#1209](https://github.com/exo-explore/exo/pull/1209))
- Less spammy logs ([#1218](https://github.com/exo-explore/exo/pull/1218), [#1225](https://github.com/exo-explore/exo/pull/1225))
- Better error messages when things go wrong ([#1198](https://github.com/exo-explore/exo/pull/1198), [#1173](https://github.com/exo-explore/exo/pull/1173), [#1177](https://github.com/exo-explore/exo/pull/1177))
- Detect Thunderbolt Bridge cycles and show a native prompt to disable Thunderbolt Bridge ([#1222](https://github.com/exo-explore/exo/pull/1222))
- Add the ability to filter nodes for placement in the dashboard by clicking them ([#1248](https://github.com/exo-explore/exo/pull/1248))

##### Bug Fixes

- Fix placement edge cases for heterogeneous devices with lopsided memory availabilities ([#1200](https://github.com/exo-explore/exo/pull/1200))
- Fix various issues with auto parallel for loading and sharding models ([#1202](https://github.com/exo-explore/exo/pull/1202), [#1201](https://github.com/exo-explore/exo/pull/1201), [#1206](https://github.com/exo-explore/exo/pull/1206), [#1211](https://github.com/exo-explore/exo/pull/1211), [#1229](https://github.com/exo-explore/exo/pull/1229), [#1223](https://github.com/exo-explore/exo/pull/1223))
- Prepend think tags for certain models that include a think tag in their chat template e.g. GLM 4.7 ([#1186](https://github.com/exo-explore/exo/pull/1186))

**Full Changelog:** https://github.com/exo-explore/exo/compare/v1.0.63...v1.0.64

### v1.0.63

- Date: 2026-01-16
- Version: v1.0.63
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.63
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.63

#### EXO v1.0.63 Release Notes

This release comes with stability improvements for long-running clusters, and support for GLM 4.7 and MiniMax M2.1.

##### Model Support

- Added support for GLM 4.7 (#1147)
- Added support for MiniMax M2.1 (#1147)
- Tensor sharding for GPT-OSS (#1144)

##### UI

- Uninstall button in macOS app (#1077)

##### Bug Fixes

- Fix issues with nodes incorrectly dropping out of topology (#1164, #1170)
- Fix `exo-bench` for transformers 5.x (#1168)
- Fix gibberish responses for GPT-OSS (#1165)
- Increase rlimit to avoid hitting resource limit errors when launching more than 10 instances (#1148)
- Reduce false-positives on local network access warning in macOS app (#1136)
- Use user-provided `seed` for sampling in generation, enabling deterministic responses (#1094)
- Change status shown in Dashboard/macOS app from `UNKNOWN` to `PREPARING` (#1112)
- Cleaner scrolling in instances box on Dashboard (#1113)

**Full Changelog:** https://github.com/exo-explore/exo/compare/v1.0.62...v1.0.63

### v1.0.62

- Date: 2026-01-08
- Version: v1.0.62
- Original notes: https://github.com/exo-explore/exo/releases/tag/v1.0.62
- Permalink: https://whatsnew.fyi/product/exo/releases/v1.0.62

##### What's Changed
* ci: avoid uploading alpha appcasts by @JakeHillion in https://github.com/exo-explore/exo/pull/1110
* ci: compute CURRENT_PROJECT_VERSION from semver by @JakeHillion in https://github.com/exo-explore/exo/pull/1111


**Full Changelog**: https://github.com/exo-explore/exo/compare/v1.0.61...v1.0.62