Ollama app icon

Ollama Preview Archive

Older signed macOS MLX preview builds kept for comparison, rollback testing, and release provenance. Current builds remain on the main download page.

Ollama 0.34.1 MLX preview

Published September 18, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.34.1-30-gc8b964e-mlx-preview-v0.34.1.dmg
Size
192.5 MB
SHA-256
bf0da003659be88682e1605628364d559c2cddffabae6f52f18cde57ae94f845
Upstream
Ollama v0.34.1 @ 38fdb5dd58c7
Highlights
v0.34.1 base with GPT-OSS, Apertus, Gemma 4 media, Responses tools, and desktop audio attachments. Official Ollama v0.34.1Updated official baseline with the current app, API, model-create, and runtime behavior from the stable v0.34.1 release. Included MLX preview capabilitiesNative GPT-OSS runtime with validated thinking and tool calling; Apertus 1.0 text and Apertus 1.1 Mini runtimes (see the model caveats)
Gemma 4 and Apertus 1.5 Omni image/audio with MP3, WebP, cancellation, and bounded media processing
Responses namespace tools, custom apply_patch compatibility, sparse tokenizer allocation, and desktop audio attachments.
Download DMG

Ollama 0.33.3 MLX preview

Published September 5, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.33.3-75-g9581bea-mlx-preview-v0.33.3.dmg
Size
155.9 MB
SHA-256
529fd4c2fc81109e53b8e94db88a74ff6d339f10b1758d748bcaa7c7ab160063
Upstream
Ollama v0.33.3 @ b79067b0db74
Highlights
v0.33.3 base with GPT-OSS, Apertus, Gemma 4 media, Responses tools, and desktop audio attachments. Official Ollama v0.33.3New since v0.33.2 preview: Gemma 4 image/audio support on MLX, cached prompt-token reporting, model-defined GGUF defaults, and updated MLX, MLX-C, and llama.cpp. Included MLX preview capabilitiesNative GPT-OSS and Apertus runtimes, including retained Apertus Mini 4B behavior
Gemma 4 and Apertus 1.5 Omni image/audio with MP3, WebP, and bounded media handling
Thinking, tool calling, Responses namespaces, custom apply_patch, and desktop audio attachments across validated preview models.
Download DMG

Ollama 0.33.2 MLX preview

Published August 31, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.33.2-90-g6c08853-mlx-preview-v0.33.2.dmg
Size
155.9 MB
SHA-256
37fbe56a247acb85cadab01422e7f625544195f056a8dfff7a31f9555bc83422
Upstream
Ollama v0.33.2 @ f96e7aa0513b
Highlights
Official Ollama v0.33.2New since v0.33.1 preview: updated official baseline with desktop and Claude-proxy behavior refinements plus discovery and ML API cleanup. Included MLX preview capabilitiesNative GPT-OSS and Apertus runtimes, including retained Apertus Mini 4B behavior
Gemma 4 and Apertus 1.5 Omni ordered image/audio with bounded media handling
Thinking, tool calling, Responses namespaces, custom apply_patch, and desktop audio attachments across validated preview models. Release processThis first refresh used two collaborating AI agents to reconstruct and rebase the PR stack step by step, then validate functionality mostly autonomously. The user retained approval, signing, and publication decisions. Performance evidence is thermally inconclusive, so no quantitative performance claim is made.
Download DMG

Ollama 0.33.1 MLX preview

Published August 28, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.33.1-96-g6ffd70f-mlx-preview-v0.33.1.dmg
Size
155.9 MB
SHA-256
c18edd7ddff69027dac052b396d83bc2715c90c32985743113206cdb963c7841
Upstream
Ollama v0.33.1 @ 13f2fb8c9927
Highlights
Official Ollama v0.33.1 New since v0.32.15 preview: Claude Desktop integration, more reliable resumed-prefill caching, structured MLX output, Qwen3.8 Flash Next, updated MLX and llama.cpp runtimes, and safer loading from slow storage. Included MLX preview capabilities Native GPT-OSS and Apertus runtimes, including Apertus 1.1 Mini affine imports
Gemma 4 and Apertus 1.5 Omni ordered image/audio with quantized media, MP3, WebP, cancellation, and bounded processing
Thinking, tool calling, Responses namespaces, and custom apply_patch support across validated preview models.
Download DMG

Ollama 0.32.15 MLX preview — r2

Published August 23, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.32.15-78-g5b266ac-mlx-preview-v0.32.15-r2.dmg
Size
150.1 MB
SHA-256
6ed699b80b8b2babe0082ee76c00a5fcb02926fdadac182858e8f5c0f7c3b809
Upstream
Ollama v0.32.15 @ b7871fc0d1d8
Highlights
v0.32.15 r2 with native GPT-OSS MXFP4 preservation, packed Apertus 70B MXFP8, media models, and Responses tools. Official Ollama v0.32.15 New since v0.32.13 preview: desktop onboarding, safer mid-stream parser failures, WebP transcoding for llama-server, more tolerant Qwen system-message placement, and cached model metadata. Included MLX preview capabilities Native GPT-OSS and Apertus runtimes, including Apertus 1.1 Mini imports and published artifacts under conversation-quality investigation
Gemma 4 and Apertus 1.5 Omni image/audio with MP3, WebP, bounded media handling, and desktop audio attachments
Thinking, tool calling, and Responses namespace/custom apply_patch support across validated preview models. r2 release fixes Preserves native GPT-OSS MXFP4 snapshots and imports Apertus 70B as packed MXFP8. The signed DMG freshly created and ran GPT-OSS 20B/120B plus Apertus 70B.
Download DMG

Ollama 0.32.15 MLX preview

Published August 23, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.32.15-72-g4ee3121-mlx-preview-v0.32.15.dmg — superseded; no download link published here
Size
150.1 MB
Superseded by v0.32.15-r2
SHA-256
9a5247c04b27f09973dcda7c61fd7a2a3734a096e08895d107b759406e9182c1
Upstream
Ollama v0.32.15 @ b7871fc0d1d8
Disposition
Replaced the same day by r2, which preserves native GPT-OSS MXFP4 snapshots and imports Apertus 70B as packed MXFP8.

Ollama 0.32.13 MLX preview — Codex apply_patch update

Published August 22, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.32.13-74-gcb9b2ae-mlx-preview-v0.32.13-applypatch1.dmg
Size
150.2 MB
SHA-256
304a403bb320e22e9707aaf086df68daebfa57477a5501aea440b672644b2952
Upstream
Ollama v0.32.13 @ 0f25c31bd53b
Highlights
v0.32.13 base with desktop audio attachments, Responses apply_patch compatibility, and the complete MLX model stack. Official Ollama v0.32.13 Qwen 3.8 27B with Apple Silicon MLX optimization and developer-instruction support. Included MLX preview capabilities Native GPT-OSS and Apertus runtimes, including Apertus 1.1 Mini INT2–INT8 imports
Gemma 4 and Apertus 1.5 Omni image/audio with MP3, WebP, and bounded media handling
Thinking and tool calling across the supported preview models. Desktop audio attachments Select or drop MP3 and WAV files in chat; supported audio models receive the media bytes through the established chat transport. Codex apply_patch compatibility Exposes Codex custom apply_patch tools through the Responses API, while keeping ordinary non-GPT-OSS models on their supported tool path.
Download DMG

Ollama 0.32.13 MLX preview — audio attachment update

Published August 22, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.32.13-71-gf626368-mlx-preview-v0.32.13-audiofix1.dmg — superseded; no download link published here
Size
150.2 MB
Superseded by apply_patch update
SHA-256
5a13a5c8b69c83f1a5345390ec40e42c2a5888d7bf70150b2952a3bb6284bac9
Upstream
Ollama v0.32.13 @ 0f25c31bd53b
Disposition
Superseded the same day by the complete v0.32.13 apply_patch compatibility update.
Highlights
Official Ollama v0.32.13 New since v0.32.11 preview: Qwen 3.8 27B with Apple Silicon MLX optimization and developer-instruction support Included MLX preview capabilities Native GPT-OSS and Apertus runtimes, Apertus 1.1 Mini imports, Gemma 4 and Apertus 1.5 Omni media, thinking, tools, and Codex-compatible Responses support Desktop audio attachments Select or drop MP3 and WAV files in chat; supported audio models receive the media bytes through the established chat transport.

Ollama 0.32.11 MLX preview

Published August 14, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.32.11-69-g347279a-mlx-preview-v0.32.11.dmg
Size
150.1 MB
SHA-256
780ba04c5a8e82b88c1160954db00cdc60c95f55ea275f2aedba78d570a45c2e
Upstream
Ollama v0.32.11 @ 39df91c9826b
Highlights
v0.32.11 base with GPT-OSS, Apertus 1.1 Mini, Gemma 4 and Apertus 1.5 Omni media, and Responses tools. Official Ollama v0.32.11 New since v0.32.9 preview: Responses API web search, DeepSeek Harness and Muse Code launch integrations, faster global-scale NVFP4 MLX prefill, and Muse Glimmer template updates Included MLX preview capabilities Native GPT-OSS and Apertus runtimes, including Apertus 1.1 Mini INT2–INT8 imports
Gemma 4 and Apertus 1.5 Omni image/audio with MP3, WebP, and bounded media handling
Thinking, tool calling, and Codex-compatible Responses namespaces and custom apply_patch
Download DMG

Ollama 0.32.9 MLX preview

Published August 12, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.32.9-69-g968a221-mlx-preview-v0.32.9.dmg
Size
144.7 MB
SHA-256
fce34fdbb0fc32fcfdfbb7c65a4c561e168280eae9bcd970be957a7ae6161e18
Upstream
Ollama v0.32.9 @ 1d5febee105f
Highlights
v0.32.9 base with GPT-OSS, Apertus 1.1 Mini, Gemma 4 and Apertus 1.5 Omni media, and Responses tools. Official Ollama v0.32.9 New since v0.32.6 preview: Muse Glimmer with MLX, DFlash, image input, and coding-agent support, plus Nemotron 3.5 Lightning Included MLX preview capabilities Native GPT-OSS and Apertus runtimes, including Apertus 1.1 Mini INT2–INT8 imports
Gemma 4 and Apertus 1.5 Omni image/audio with MP3, WebP, and bounded media handling
Thinking, tool calling, and Codex-compatible Responses namespaces and custom apply_patch
Download DMG

Ollama 0.32.6 MLX preview

Published August 8, 2026 · signed and notarized macOS DMG
Archived
File
Ollama-0.32.6-71-gfb4ced9-mlx-preview-v0.32.6.dmg
Size
142.7 MB
SHA-256
7ad6dec5f4a9ee13ee3537fdf159025b47e871244a9151009a268663734b0604
Upstream
Ollama v0.32.6 @ c82ebbd5bfb9
Highlights
v0.32.6 base with GPT-OSS, Apertus 1.1 Mini, Gemma 4 and Apertus 1.5 Omni media, and Responses tools. Official Ollama v0.32.6 New since v0.31.2 preview: including agent workflows, OpenAI streaming fixes, and MLX reliability/performance updates for Qwen3.5, recurrent models, and Laguna Included MLX preview capabilities Native GPT-OSS and Apertus runtimes, including Apertus 1.1 Mini INT2–INT8 imports
Gemma 4 and Apertus 1.5 Omni image/audio with MP3, WebP, and bounded media handling
Thinking, tool calling, and Codex-compatible Responses namespaces and custom apply_patch
Download DMG

Ollama 0.31.2 MLX preview

Published August 8, 2026 · signed macOS DMG
Archived
File
Ollama-0.31.2-67-g0d2a68f-mlx-preview-v0.31.2-apertus15.dmg
Size
127.0 MB
SHA-256
61b8b9345fac6facfe2462d559fbac8ca813858ed40b33909433cb55ac00f025
Upstream
Ollama v0.31.2
Highlights
v0.31.2 base with Gemma 4, Apertus 1.5 Omni, GPT-OSS, and Responses tools. Official Ollama v0.31.2 New since v0.31.2-rc2 preview: final release tag; source commit unchanged from the rc2 preview Included MLX preview capabilities Gemma 4 unified MLX image and audio support
Apertus 1.5 Omni text, image, audio, and thinking controls
GPT-OSS MLX and Responses namespace/apply_patch tools
Gemma 4 tool-call safety integration
Download DMG

Ollama 0.31.2 MLX source preview

Prepared August 1, 2026 · Gemma 4 + Apertus 1.5 Omni · source reference
Source only
Status
Source-only reference for local builds and source artifact downloads. The signed DMG above supersedes this source preview for normal installation.
Upstream
Ollama v0.31.2
Highlights
Source reference for Gemma 4, Apertus 1.5 Omni, GPT-OSS, and Responses tools. Official Ollama v0.31.2 New since v0.31.2-rc2 preview: final release tag; source reference kept for local builds Included source preview capabilities Gemma 4 unified MLX image and audio support
Apertus 1.5 Omni text, image, audio, and thinking controls
GPT-OSS MLX and Responses namespace/apply_patch tools
Gemma 4 tool-call safety integration
Source archive only

Ollama 0.31.2-rc2 MLX preview

Published July 8, 2026 · signed macOS DMG
Archived
File
Ollama-0.31.2-rc2-13-gdf00506-mlx-preview-v0.31.2.dmg
Size
126.6 MB
Download DMG
SHA-256
851b49641b7e7e8b87d024c6f8c7395309a9ae99e285c56d5b2dcee346122208
Highlights
Official Ollama v0.31.2-rc2 New since v0.31.1 preview: x/create rewrite, agent harness work, GGUF create hardening, thinking-parser formatting fixes, iGPU projector offload padding, and MLX/llama.cpp updates Included MLX preview capabilities GPT-OSS, Apertus, namespace tools, and custom apply_patch streams

Ollama 0.31.1 MLX preview

Published July 3, 2026 · signed macOS DMG
Archived
File
Ollama-0.31.1-76-gae08bbf-mlx-preview-v0.31.1.dmg
Size
126.7 MB
Download DMG
SHA-256
db4179aeb464e6d8c90111b6b6f704ab6b19aa646e99a50235143aa1fc41f1e8
Highlights
Official Ollama v0.31.1 New since July 2 v0.31.1 preview: official base unchanged Included MLX preview capabilities GPT-OSS, Apertus, namespace tools, custom apply_patch, and MLX fix streams
Feature streams
GPT-OSS MLX @ 7133925 / 60 commits / patch-set d13ced4383
Apertus MLX @ 42f6bf5 / 5 commits / patch-set 140a9a73c2
namespace tools @ eb608e4 / 1 commit / patch-set 888ea3bacb
apply_patch tool @ a7cb2b3 / 1 commit / patch-set fe3c6cfd8a
Fix streams
MLX create worker stream fix @ 0e74971 / 1 commit / patch-set 29b27915f0
Qwen MLX MTP import skip @ f6d4676 / 1 commit / patch-set 3111ed0a2a
Integration
gofumpt return style @ acbd586 / 1 commit / patch-set d02fc440de

Ollama 0.31.1 MLX preview

Published July 2, 2026 · signed macOS DMG
Archived
File
Ollama-0.31.1-74-gacbd586-mlx-preview-v0.31.1.dmg
Size
126.7 MB
Superseded by July 3 build
SHA-256
173444edf632222e21dc3a34fe3d3c8ac5e2ab34c4c790e813c3b78a203e8742
Highlights
Official Ollama v0.31.1 New since v0.30.10 preview: Gemma 4 MoE loading fixes, MLX context-size reporting, native chat-template generation alignment, speculative-decoding runtime work, and llama.cpp/MLX dependency updates Included MLX preview capabilities GPT-OSS, Apertus, namespace tools, custom apply_patch, and MLX fix streams
Feature streams
GPT-OSS MLX @ 7133925 / 60 commits / patch-set d13ced4383
Apertus MLX @ 42f6bf5 / 5 commits / patch-set 140a9a73c2
namespace tools @ eb608e4 / 1 commit / patch-set 888ea3bacb
apply_patch tool @ a7cb2b3 / 1 commit / patch-set fe3c6cfd8a
Fix streams
MLX create worker stream fix @ 0e74971 / 1 commit / patch-set 29b27915f0
Integration
gofumpt return style @ acbd586 / 1 commit / patch-set d02fc440de

Ollama 0.30.10 MLX preview

Published June 22, 2026 · signed macOS DMG
Archived
File
Ollama-0.30.10-76-ga256e86-mlx-preview-v0.30.10.dmg
Size
140.5 MB
Download DMG
SHA-256
60e3a9c9e8d4edde8b749ea290b9f338411a2edb1de3dfcef5246781adf46fb3
Highlights
Official Ollama v0.30.10 New since June 21 v0.30.10 preview: official base unchanged Included MLX preview capabilities GPT-OSS, Apertus, namespace tools, custom apply_patch, and MLX cache/create fixes
Feature streams
GPT-OSS MLX @ 10cad87 / 58 commits / patch-set 42ebc4d6b8
Apertus MLX @ 42f6bf5 / 5 commits / patch-set 140a9a73c2
namespace tools @ eb608e4 / 1 commit / patch-set 888ea3bacb
apply_patch tool @ a7cb2b3 / 1 commit / patch-set fe3c6cfd8a
Fix streams
MLX cache restore correctness @ 85b7cd9 / 1 commit / patch-set f3c66004e6
MLX create worker stream fix @ 0e74971 / 1 commit / patch-set 29b27915f0
Integration
feature and fix stream merge integration only; no separate integration commits

Ollama 0.30.10 MLX preview

Published June 21, 2026 · signed macOS DMG
Archived
File
Ollama-0.30.10-68-g6ce4460-new-mlx-models-upstream-v0.30.10-apply-patch.dmg
Size
140.5 MB
Download DMG
SHA-256
22cf3f303fe3bcb87c7cad9c417c977b728494ff88aa9c71cc28252f7a501a33
Highlights
Official Ollama v0.30.10 New since June 19 v0.30.10 preview: official base unchanged Included MLX preview capabilities GPT-OSS, Apertus, namespace tools, custom apply_patch, and Cohere import compatibility
Feature streams
GPT-OSS MLX @ 85166f5 / 51 commits / patch-set 15bca8ffed
Apertus MLX @ d46497b / 5 commits / patch-set de950e4219
namespace tools @ cd9d5f2 / 1 commit / patch-set 888ea3bacb
apply_patch tool @ 0ef6086 / 1 commit / patch-set fe3c6cfd8a
Integration
our v0.30.10 integration fix: Cohere import transform compatibility @ 6ce4460 / 1 commit / patch-set e6a65c244e

Ollama 0.30.10 MLX preview

Published June 19, 2026 · signed macOS DMG
Archived
File
June 19 artifact superseded; no download link published here
Size
140.5 MB
Superseded by June 21 build
SHA-256
58e87875c4bd318ee1725973738b65734db61cff5594de746e6ce045fd4dd42a for the superseded June 19 artifact
Highlights
Official Ollama v0.30.10 New since v0.30.8 preview: Cohere2 MoE MLX support, context-shift handling for large context windows, context-limit errors, LFM2 thinking-template updates, and llama.cpp updates Included MLX preview capabilities GPT-OSS, Apertus, namespace tools, custom apply_patch, and Cohere import compatibility
Feature streams
GPT-OSS MLX @ 85166f5 / 51 commits / patch-set 15bca8ffed
Apertus MLX @ d46497b / 5 commits / patch-set de950e4219
namespace tools @ cd9d5f2 / 1 commit / patch-set 888ea3bacb
apply_patch tool @ 0ef6086 / 1 commit / patch-set fe3c6cfd8a
Integration
our v0.30.10 integration fix: Cohere import transform compatibility @ 6ce4460 / 1 commit / patch-set e6a65c244e

Ollama 0.30.8 MLX preview

Published June 14, 2026 · signed macOS DMG
Archived
File
Ollama-0.30.8-64-gdb13181-new-mlx-models-upstream-v0.30.8.dmg
Size
140.9 MB
Download DMG
SHA-256
10dc57ca9d708a6b7c4451d947eefc0d4e0390399a3c570bae211fc9d4b9468c
Highlights
Official Ollama v0.30.8 New since v0.30.7 preview: launch-provider drift fixes, prompt-cache and context-shift separation, MLX layer over-promotion hardening, speculative-cache snapshots, and recurrent-state handling Included MLX preview capabilities GPT-OSS, Apertus, namespace tools, and v0.30.8 integration fixes
Feature streams
GPT-OSS MLX @ 85166f5 / 51 commits / patch-set 15bca8ffed
Apertus MLX @ d46497b / 5 commits / patch-set de950e4219
namespace tools @ cd9d5f2 / 1 commit / patch-set 888ea3bacb
Integration
our v0.30.8 integration fixes: lint and MLX stream test fixes @ db13181 / 2 commits / patch-set 45eaf579d4

Ollama 0.30.7 MLX preview

Published June 6, 2026 · signed macOS DMG
Archived
File
Ollama-0.30.7-rc0-64-g46c5359-new-mlx-models.dmg
Size
140.9 MB
Upstream
Ollama v0.30.7-rc1
Download DMG
SHA-256
b4d62632e2dd9076510178b0fbf258ce41ed3cf663d287183c508ae2f057a8bc
Highlights
Official Ollama v0.30.7-rc1 New since v0.25.0-rc0 preview: Hermes/Codex/Qwen launch integrations, Gemma 4 12B support, Laguna architecture work, projector GPU offload, NVFP4 global-scale handling, and update/markdown/SSE hardening Included MLX preview capabilities GPT-OSS, Apertus, namespace tools, and refreshed v0.30.7-rc1 integration
Feature streams
GPT-OSS MLX @ 6434a8a / 51 commits / patch-set 7150276b61
Apertus MLX @ 8fb7d6c / 5 commits / patch-set f8f6bdc79c
namespace tools @ 6c84ab2 / 1 commit / patch-set 888ea3bacb
Integration
our v0.30.7-rc1 integration fix: refreshed MLX integration @ 46c5359 / 1 commit / patch-set a78539c7b0

Ollama 0.25.0-rc0 MLX preview

Published May 22, 2026 · signed macOS DMG
Archived
File
Ollama-0.25.0-rc0-66-gfa19a15-new-mlx-models.dmg
Size
137.3 MB
Download DMG
SHA-256
8ef7cbe225ae5927c84c9a9d9b42d14a65278ef86085a52b8005822f513a42b4
Highlights
Official Ollama v0.25.0-rc0 New since v0.22.1 preview: Codex App launch work, MLX sampler rework, DFlash and Gemma 4 speculative decoding, memory tracing, imagegen/threading updates, app update hardening, and Go 1.26 Included MLX preview capabilities GPT-OSS, Apertus, and namespace tools on the v0.25.0-rc0 base
Feature streams
GPT-OSS MLX @ beaba00 / 51 commits / patch-set 7150276b61
Apertus MLX @ 62f3c3f / 5 commits / patch-set f8f6bdc79c
namespace tools @ 6c84ab2 / 1 commit / patch-set 888ea3bacb
Integration
feature merge integration only; no separate integration commits

Ollama 0.22.1 GPT-OSS MLX test

Published May 1, 2026 · macOS DMG
Archived
File
Ollama-0.22.1-gptoss-mlx-test-20260501.dmg
Size
161.5 MB
Download DMG
SHA-256
f47d5d1d4f18a408471a8da01f07dc7c634cf5c3a1ef22db9059a6ff90b3c8d3
Highlights
Official Ollama v0.22.1 First MLX preview on this archive page; no earlier preview baseline is listed here Included MLX preview capabilities Initial GPT-OSS MLX test stack
Feature streams
GPT-OSS MLX commit stack / 54 commits / patch-set dd0e74ab24
Integration
linear GPT-OSS test stack; no separate integration manifest

To verify a download, run shasum -a 256 <downloaded-file.dmg> and compare it with the SHA-256 value shown above.

AI assistance disclosure: this page's release notes and explanatory copy were edited with AI assistance and reviewed before publication.