-
ddb710507d
feat(deepseek): add support for Responses API and Chat Prefix Completion
main
hobokenchicken
2026-08-03 03:53:42 +00:00
-
2ed027da7c
fix: scope reasoning_effort=none to OpenAI provider only (Xiaomi rejects 'none')
newkirk
2026-07-30 15:26:28 -04:00
-
2e4253e0bf
fix: add ModelGroup field to LLMRequest struct matching DB schema
newkirk
2026-07-30 15:20:53 -04:00
-
2562675b8b
fix: set reasoning_effort=none when tools present for OpenAI reasoning models
newkirk
2026-07-30 15:18:40 -04:00
-
b3dc4365a8
fix(dustins_stack): correct kimi model name kimi-2.6 → kimi-k2.6
newkirk
2026-07-29 12:27:07 -04:00
-
1200a081fc
fix(dustins_stack): remove gemini-3.5-flash-lite from targets and fallbacks
newkirk
2026-07-29 12:22:30 -04:00
-
c6329d5612
fix(gemini): resolve latency and error issues with model normalization and v1beta tool routing
hobokenchicken
2026-07-28 17:58:21 +00:00
-
c96e1d0350
mobile: off-canvas sidebar, responsive grids/charts/top-bar
newkirk
2026-07-27 23:47:23 -04:00
-
ba25f5e6c8
style: optimize stat-card grid to fit on 1 line for desktop screens or 2 lines for laptops
hobokenchicken
2026-07-21 18:14:24 +00:00
-
59c38e8f8f
feat: add Days Active card and total days lifetime span to dashboard overview
hobokenchicken
2026-07-21 18:11:18 +00:00
-
d23da551ab
docs: update README with Xiaomi MiMo support, HTTP connection pooling, token caching, and thread safety
newkirk
2026-07-21 13:43:57 -04:00
-
654f6ab6d1
perf: add HTTP connection pooling across providers and disable proxy buffering on SSE streams
hobokenchicken
2026-07-21 17:40:18 +00:00
-
0decc63e8c
feat: add token caching middleware and max history messages config
hobokenchicken
2026-07-21 17:30:22 +00:00
-
9980123f97
fix: optimize Gemini stream parsing to support real-time streaming of JSON arrays
hobokenchicken
2026-07-17 18:52:54 +00:00
-
293cf057b9
fix: restrict tool-heavy/swe-bench tagging to software development or agent-related tasks
hobokenchicken
2026-07-17 17:25:17 +00:00
-
eb90f949a0
fix: Gemini provider compatibility with Gemini 3 models
hobokenchicken
2026-07-17 13:41:41 +00:00
-
a187d8e20e
fix(phase-4): add DB error checking, mask sensitive auth tokens in API and middleware logs
newkirk
2026-07-21 13:23:02 -04:00
-
700b7cd5d6
fix(phase-3): enable circuit breaker for streaming requests
newkirk
2026-07-21 13:21:39 -04:00
-
4027ed4351
fix(phase-2): add thread-safety to ModelRegistry and Router reload operations
newkirk
2026-07-21 13:20:29 -04:00
-
42b70621a1
fix(phase-1): redact API key query params from log output and add mutex thread-safety to providers map
newkirk
2026-07-21 13:18:31 -04:00
-
84a18f5866
fix(classifier): add tier boundaries and signal keywords to dispatcher prompt
newkirk
2026-07-09 15:46:27 -04:00
-
25e246061f
fix: resolve retired/preview gemini model routing and test configuration errors
hobokenchicken
2026-06-18 13:32:34 +00:00
-
73a82e6175
feat: implement advanced condition-based heuristic model routing
hobokenchicken
2026-06-05 15:05:13 +00:00
-
b3354a1bbc
Add Xiaomi MiMo provider (mimo-v2.5) support
newkirk
2026-05-29 12:19:24 -04:00
-
1dc5f586b9
fix: improve OpenAI error body capture and log request body on 400
newkirk
2026-05-17 19:57:59 -04:00
-
40f055cb57
fix: correct deepseek pricing, gemini streaming tokens, and group-name logging
newkirk
2026-05-17 19:48:47 -04:00
-
970e778703
chore: update .gitignore to ignore nohup.out and bak files
hobokenchicken
2026-05-11 03:13:54 +00:00
-
477a811999
fix: remove tool call ID truncation and improve DeepSeek reasoning handling
hobokenchicken
2026-05-11 03:12:38 +00:00
-
d2b9da89d9
fix FindModel: prioritize canonical providers to prevent reseller limit overrides
hobokenchicken
2026-05-07 14:47:17 -04:00
-
b7df3108fa
docs: update README, TODO, and deployment docs
hobokenchicken
2026-05-07 14:07:52 -04:00
-
28b8271c1d
fix: inject English system prompt for DeepSeek provider
hobokenchicken
2026-05-07 14:03:39 -04:00
-
eb585c0001
fix: switch dispatcher classifier to gpt-5.4-nano
hobokenchicken
2026-05-07 14:00:19 -04:00
-
4aea7a3b4c
fix: select provider AFTER routing resolves model groups
hobokenchicken
2026-05-07 13:54:42 -04:00
-
330eaa57d1
fix: update model names to match current models.dev registry
hobokenchicken
2026-05-07 13:48:33 -04:00
-
0ae30036f0
fix: classifier selector model now routes to correct provider
hobokenchicken
2026-05-07 13:37:19 -04:00
-
3c0b59622e
feat: classifier bucket mapping + dispatcher seed group
hobokenchicken
2026-05-07 13:18:35 -04:00
-
7517307c11
feat: add hierarchical routing — groups can target other groups
hobokenchicken
2026-05-07 12:28:31 -04:00
-
19517b0847
chore: add deploy.sh for prod restarts
hobokenchicken
2026-05-07 12:02:28 -04:00
-
a3a6f765e7
feat: add logic_level and primary_use metadata to model groups
hobokenchicken
2026-05-07 12:01:28 -04:00
-
79dd122b56
feat: expose model groups in /v1/models endpoint
hobokenchicken
2026-05-07 11:26:05 -04:00
-
3021e4b2b4
fix: log resolved model name instead of group name in Recent Activity
hobokenchicken
2026-05-07 11:16:36 -04:00
-
14de7e9ebf
fix: wrap model-groups API responses in SuccessResponse for api.js client
hobokenchicken
2026-05-05 11:41:23 -04:00
-
4fef201e95
fix: remove /api prefix from model-groups API calls (api.js already prepends it)
hobokenchicken
2026-05-05 11:33:05 -04:00
-
bac03de051
docs: add automatic model routing to README
hobokenchicken
2026-05-05 11:28:59 -04:00
-
37949e560b
feat: add model groups dashboard page with CRUD UI
hobokenchicken
2026-05-05 10:55:25 -04:00
-
f04cb6b8f2
feat: add model groups CRUD admin API endpoints
hobokenchicken
2026-05-05 10:50:33 -04:00
-
10262c0e5a
feat: wire model group router into chat completions handler
hobokenchicken
2026-05-05 10:47:32 -04:00
-
d345f8c41d
feat: add classifier routing strategy with LLM complexity rating
hobokenchicken
2026-05-05 10:40:26 -04:00
-
d1f7a57f58
feat: add router package with heuristic strategy
hobokenchicken
2026-05-05 10:37:36 -04:00
-
dc9af4d79c
feat: add model_groups table and default seed data
hobokenchicken
2026-05-05 10:33:35 -04:00
-
c009d401fb
docs: add Responses API endpoint to README
hobokenchicken
2026-05-05 09:36:51 -04:00
-
e5ef39f327
feat: add OpenAI Responses API support (POST /v1/responses)
hobokenchicken
2026-05-02 16:38:17 -04:00
-
eb67287b56
fix: raise provider HTTP timeouts from 30s to 10min
hobokenchicken
2026-04-30 10:17:45 -04:00
-
4aa17b4fd2
debug: add max_tokens trace logging to chat completions handler
hobokenchicken
2026-04-30 10:04:50 -04:00
-
79571c6bdc
fix: replace sql.NullTime with string scan for MAX() aggregate queries
hobokenchicken
2026-04-30 09:32:11 -04:00
-
d46a333249
feat: inject max_tokens from models.dev registry when not specified in request
hobokenchicken
2026-04-28 15:36:06 -04:00
-
7446f3463d
fix: add per-image cost tracking for DALL-E and Imagen
hobokenchicken
2026-04-27 10:42:29 -04:00
-
b1a72f5a10
fix: estimate image gen tokens from prompt length instead of hardcoding
hobokenchicken
2026-04-27 10:28:39 -04:00
-
5ee539d95c
feat: add image generation for OpenAI DALL-E and Gemini Imagen
hobokenchicken
2026-04-27 10:06:07 -04:00
-
14e26a4323
feat: capture Gemini cached content tokens in cost tracking
hobokenchicken
2026-04-26 21:14:53 -04:00
-
1c3b1c6fe9
fix: FindModel reverse fuzzy match for date-suffixed model IDs
hobokenchicken
2026-04-26 21:09:56 -04:00
-
5e0c10db01
fix: goimports — strip unused imports from all server files
hobokenchicken
2026-04-26 15:00:04 -04:00
-
e598150d90
fix: trim imports per file — no unused imports in split files
hobokenchicken
2026-04-26 14:58:53 -04:00
-
2fa6f0df62
fix: split dashboard.go properly — extract analytics + models_config
hobokenchicken
2026-04-26 14:57:28 -04:00
-
db76858072
fix: import block syntax in split dashboard files
hobokenchicken
2026-04-26 14:55:29 -04:00
-
af2c5b95f7
feat: Phase 3 - architecture & maintainability
hobokenchicken
2026-04-26 14:52:10 -04:00
-
1f574d8134
feat: Phase 2 - reliability & observability
hobokenchicken
2026-04-26 14:48:41 -04:00
-
8a8d8d1477
fix: Phase 1 - security & stability patches
hobokenchicken
2026-04-26 14:45:22 -04:00
-
da074f52b4
fix: remove global auth middleware, causing webui login issues
hobokenchicken
2026-04-09 12:21:02 -04:00
-
9b0aa4dbe8
fix: remove unused fmt import in circuit breaker provider
hobokenchicken
2026-04-09 12:19:14 -04:00
-
212ac14a1b
feat: implement circuit breaker, fix auth vulnerability
hobokenchicken
2026-04-09 12:17:18 -04:00
-
2929f51556
fixed model visibility
hobokenchicken
2026-04-09 12:13:53 +00:00
-
e12418cc4c
fix(gemini): ensure strict 1:1 pairing of model calls and function responses
hobokenchicken
2026-04-07 18:57:13 +00:00
-
be4ec3482a
fix(gemini): group adjacent tool messages and ensure correct role sequence
hobokenchicken
2026-04-07 18:50:48 +00:00
-
e67aafdac1
fix(gemini): improve tool-calling support and handle function_call response
hobokenchicken
2026-04-07 18:37:57 +00:00
-
21e5204abd
fix(ollama): improve tool-calling support and restore gemma/llama context limits
hobokenchicken
2026-04-07 14:24:23 +00:00
-
4095c68822
fix(ollama): improve model detection and ensure robust token/context limits
hobokenchicken
2026-04-07 14:05:21 +00:00
-
ef37dc5af0
fix(ollama): significantly increase context and prediction limits
hobokenchicken
2026-04-07 13:48:02 +00:00
-
fdbb068a6c
fix(ollama): map max_tokens to num_predict and increase context window
hobokenchicken
2026-04-07 13:44:17 +00:00
-
dbbf48cb14
fix(ollama): increase timeout and add default max_tokens for large models
hobokenchicken
2026-04-07 13:42:10 +00:00
-
1e13b0376b
feat(ollama): improve configuration and dashboard integration
hobokenchicken
2026-04-07 12:53:17 +00:00
-
1b5cd2815e
fix(ollama): improve error handling and add timeouts
hobokenchicken
2026-04-06 15:05:31 -04:00
-
ba4c4af2f8
docs: update documentation for Ollama provider
hobokenchicken
2026-04-06 15:01:55 -04:00
-
e56a284415
docs: update TODO.md with Ollama provider completion
hobokenchicken
2026-04-06 14:46:32 -04:00
-
cbc9eeb453
fix(server): add Ollama model detection and registry support
hobokenchicken
2026-04-06 14:45:57 -04:00
-
2f6b7deb2c
feat(providers): add Ollama provider support
hobokenchicken
2026-04-06 14:38:35 -04:00
-
9375448087
fix(moonshot): resolve 401 Unauthorized errors by trimming API keys and improving request compatibility
hobokenchicken
2026-03-26 17:09:27 +00:00
-
5be2f6f7aa
fix: use Moonshot test model
hobokenchicken
2026-03-26 10:12:44 -04:00
-
eebcadcba1
fix: surface moonshot on providers page
hobokenchicken
2026-03-25 09:35:08 -04:00
-
6b2bd13903
chore: remove tracked binary gophergate
hobokenchicken
2026-03-25 13:32:51 +00:00
-
5dfda0a10c
merge: resolve conflicts in server.go and integrate moonshot support
hobokenchicken
2026-03-25 13:32:40 +00:00
-
-
a8a02d9e1c
feat: add moonshot kimi k2.5 support
hobokenchicken
2026-03-25 09:26:53 -04:00
-
bd1d17cc4d
feat: add moonshot kimi k2.5 support
hobokenchicken
2026-03-25 09:26:53 -04:00
-
-
9207a7231c
chore: update all grok-2 references to grok-4
hobokenchicken
2026-03-25 13:17:06 +00:00
-
c6efff9034
fix: update grok test model to grok-4-1-fast-non-reasoning
hobokenchicken
2026-03-25 13:14:31 +00:00
-
27fbd8ed15
chore: cleanup repository and update gitignore
hobokenchicken
2026-03-25 13:08:33 +00:00
-
348341f304
fix: prioritize database provider configs and implement API key encryption
hobokenchicken
2026-03-25 13:04:26 +00:00
-
9380580504
fix: resolve dashboard websocket 'disconnected' status
hobokenchicken
2026-03-19 14:32:34 -04:00
-
08cf5cc1d9
fix: improve cost tracking accuracy for modern models
hobokenchicken
2026-03-19 14:14:54 -04:00
-
0f0486d8d4
fix: resolve user dashboard field mapping and session consistency
hobokenchicken
2026-03-19 14:01:59 -04:00