Faisal Amir
0e972646e8
Merge pull request #6465 from menloresearch/fix/attachment-edit-message
...
fix: attachment edit message
2025-09-16 11:17:17 +07:00
Faisal Amir
4ef64ce18c
fix: linter
2025-09-16 11:16:53 +07:00
Faisal Amir
9fe517dbf8
fix: immediate dropdown value update
2025-09-16 11:16:35 +07:00
Louis
6dde7e0ba1
Merge pull request #6473 from theishangoswami/exa-mcp-addition
...
Added Exa MCP
2025-09-16 10:28:44 +07:00
Faisal Amir
da2f790f25
Merge pull request #6464 from menloresearch/fix/pathname-install-be
...
fix: pathname file install BE
2025-09-16 10:14:02 +07:00
theishangoswami
c02d8200ac
added exa mcp
2025-09-16 04:38:03 +05:30
Dinh Long Nguyen
491012fa87
remove assistant from web ( #6468 )
2025-09-15 23:53:59 +07:00
Faisal Amir
52fe8e8c0e
fix: type imageurl
2025-09-15 21:58:24 +07:00
Faisal Amir
7b9b9666cb
fix: imporove edit message with attachment image
2025-09-15 21:48:19 +07:00
Faisal Amir
3b22f0b7c0
fix: imporove edit message with attachment image
2025-09-15 21:48:01 +07:00
Louis
cf87313f28
Merge pull request #6384 from maxx-ukoo/mk_add_configurable_timeout_to_local_api_server
...
Add model response timeout for local api server as configurable value
2025-09-15 21:26:07 +07:00
Faisal Amir
e02be47aae
fix: remove log
2025-09-15 21:09:08 +07:00
Faisal Amir
5736d7b110
fix: auto update should not block popup
2025-09-15 20:51:27 +07:00
Akarshan Biswas
9e3a77a559
fix: set default memory mode and clean up unused import ( #6463 )
...
Use fallback value 'high' for memory_util config and remove unused GgufMetadata import.
2025-09-15 19:00:46 +05:30
Louis
e78e4e5cca
Merge pull request #6278 from lugnicca/feat/model-selector
...
feat: add model selector (fetch from v1/models) when user adds a provider model
2025-09-15 20:23:22 +07:00
Dinh Long Nguyen
311a451005
Always allow MCP for web ( #6462 )
...
* mcp and extension setting disabled + always allow mcp tools on web
* fix tests
2025-09-15 20:13:46 +07:00
Maksym Krasovakyi
71e2e24112
Add model response timeout for local api server as configurable value via UI
2025-09-15 14:25:09 +03:00
Faisal Amir
18114c0a15
fix: pathname file install BE
2025-09-15 18:05:11 +07:00
Bui Quang Huy
59b2755810
Merge pull request #6461 from menloresearch/docs/missing-url
...
chore: missing local api server
2025-09-15 18:16:37 +08:00
Trang Le
1aa39392ab
set bullet list style to be circle ( #6437 )
2025-09-15 17:13:07 +07:00
Faisal Amir
7921bb21c7
chore: missing local api server
2025-09-15 15:40:46 +07:00
Faisal Amir
d73a932773
Merge pull request #6460 from menloresearch/revert/old-docs
...
chore: revert back old docs
2025-09-15 15:28:14 +07:00
Faisal Amir
1ca765476e
chore: revert back old docs
2025-09-15 15:22:00 +07:00
Akarshan Biswas
e80a865def
fix: detect allocation failures as out-of-memory errors ( #6459 )
...
The Llama.cpp backend can emit the phrase “failed to allocate” when it runs out of memory.
Adding this check ensures such messages are correctly classified as out‑of‑memory errors,
providing more accurate error handling CPU backends.
2025-09-15 12:35:24 +05:30
Bui Quang Huy
32ff0df6c9
Merge pull request #6451 from menloresearch/LazyYuuki-patch-1
...
Update 4-goal.md
2025-09-15 13:07:32 +08:00
Bui Quang Huy
f02f001266
Merge branch 'dev' into LazyYuuki-patch-1
2025-09-15 13:07:11 +08:00
Nguyen Ngoc Minh
55edc7129e
Merge pull request #6457 from menloresearch/chore/makefile-rust-targets
...
chore: makefile rust targets macos
2025-09-15 12:03:17 +07:00
Louis
43431c26e7
Merge branch 'dev' into feat/model-selector
2025-09-15 12:02:25 +07:00
Akarshan Biswas
489c5a3d9c
fix: KVCache size calculation and refactor ( #6438 )
...
- Removed the unused `getKVCachePerToken` helper and replaced it with a unified `estimateKVCache` that returns both total size and per‑token size.
- Fixed the KV cache size calculation to account for all layers, correcting previous under‑estimation.
- Added proper clamping of user‑requested context lengths to the model’s maximum.
- Refactored VRAM budgeting: introduced explicit reserves, fixed engine overhead, and separate multipliers for VRAM and system RAM based on memory mode.
- Implemented a more robust planning flow with clear GPU, Hybrid, and CPU pathways, including fallback configurations when resources are insufficient.
- Updated default context length handling and safety buffers to prevent OOM situations.
- Adjusted usable memory percentage to 90 % and refined logging for easier debugging.
2025-09-15 10:16:13 +05:30
Minh141120
1db67ea9a2
chore: simplify macos workflow
2025-09-15 11:24:11 +07:00
Faisal Amir
7a2782e6fd
Merge pull request #6456 from menloresearch/enhancement/copy-mcp-permission
...
enhancement: copy MCP permission
2025-09-15 11:01:09 +07:00
Faisal Amir
a4483b7eb7
Update web-app/src/locales/en/tool-approval.json
...
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
2025-09-15 10:38:36 +07:00
Faisal Amir
a26445e557
chore: make action mutton capitalize
2025-09-15 10:34:27 +07:00
Faisal Amir
44893bc3c3
enhancement: copy MCP permission
2025-09-15 10:33:05 +07:00
Louis
1d36a4ad6e
refactor: clean up empty folders ( #6454 )
2025-09-15 10:27:07 +07:00
Minh141120
4fa78fa892
fix: make install-rust-targets a dependency
2025-09-15 09:44:21 +07:00
Minh141120
6959329fd6
chore: add install-rust-targets step for macOS universal builds
2025-09-15 09:32:51 +07:00
Dinh Long Nguyen
0771b998a5
Fix: Web Services Improvement
...
Fix: Web Services Improvement
2025-09-15 09:08:30 +07:00
Bui Quang Huy
78d816b435
Update 4-goal.md
2025-09-14 22:26:57 +08:00
Akarshan Biswas
654e566dcb
fix: correct context shift flag handling in LlamaCPP extension ( #6404 ) ( #6431 )
...
* fix: correct context shift flag handling in LlamaCPP extension
The previous implementation added the `--no-context-shift` flag when `cfg.ctx_shift` was disabled, which conflicted with the llama.cpp CLI where the presence of `--context-shift` enables the feature.
The logic is updated to push `--context-shift` only when `cfg.ctx_shift` is true, ensuring the extension passes the correct argument and behaves as expected.
* feat: detect model out of context during generation
---------
Co-authored-by: Dinh Long Nguyen <dinhlongviolin1@gmail.com>
2025-09-12 13:43:31 +05:30
Nguyen Ngoc Minh
596930174f
Merge pull request #6433 from menloresearch/ci/jan-web
...
ci: fix typo in branch name for jan web
2025-09-12 14:16:29 +07:00
Nguyen Ngoc Minh
b8ca6f5d76
ci: fix typo in branch name for jan web
2025-09-12 14:15:27 +07:00
Nguyen Ngoc Minh
91e17dc4ee
Merge pull request #6430 from menloresearch/ci/remove-path-for-jan-web
...
ci: remove paths triggered for jan server
2025-09-12 13:44:27 +07:00
Minh141120
6c935132a9
ci: remove paths triggered for jan server
2025-09-12 13:41:48 +07:00
Piotr Orzechowski
129c64a9a4
fix: Polish translation ( #6421 )
2025-09-12 12:53:58 +07:00
Faisal Amir
ad428f587b
Merge pull request #6426 from menloresearch/fix/error-validate-nested-dom
...
fix: avoid error validate nested DOM
2025-09-12 12:28:46 +07:00
Dinh Long Nguyen
b5b6e1dc19
add mcp for web ( #6411 )
...
* add mcp for web
* update /jan/v1 endpoint to /v1
* update mise and makefile
* update yarn lock
* use mcp oauth properly
2025-09-12 12:14:10 +07:00
Nguyen Ngoc Minh
72128117a9
Merge pull request #6427 from menloresearch/ci/separate-jan-web-build-and-deploy
...
ci: separate jan web build and deploy
2025-09-12 11:20:09 +07:00
Minh141120
32440a9510
ci: separate jan web build and deploy
2025-09-12 11:10:28 +07:00
Faisal Amir
4293fe7edc
fix: avoid error validate nested dom
2025-09-12 10:58:34 +07:00
