scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-24 00:52:24 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 90
50181 commits 2663 branches 1420 tags 3.4 GiB
Commit graph

50181 commits

This branch
This branch
All branches
Author SHA1 Message Date
yassin
7095373dd5 fix(proxy): only discard parked stream logging for errors the failure path logs 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 20:10:19 +00:00
jesus
ec799686a4 style(proxy): ruff format utils.py 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 20:10:19 +00:00
jesus
0949f24eef refactor(proxy): tighten deferred stream logging discard docstring 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 20:10:18 +00:00
jesus
8b24d4c24f fix(proxy): log blocked streaming guardrail responses as failures, not success 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 20:10:18 +00:00
Devin AI
abd1ea1b1c test(proxy): trim reset budget race test comments 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 20:06:29 +00:00
yassin
2505fcf95e test(proxy): type the fixture parameters of the org member redis test 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 20:05:23 +00:00
tin-berri
d07b2e87d2
Merge pull request #41270 from BerriAI/litellm_capability_classifier_pr 
feat(router): add capability classifier as Fuse foundation
 2026-09-15 13:04:21 -07:00
Yassin Kortam
435f0d22e8
Merge pull request #41156 from BerriAI/litellm_weighted_routing_model_id 
fix(router): keep weighted routing when a deployment id equals a model_name
 2026-09-15 12:55:45 -07:00
yassin
264305de23 fix(proxy): keep access-group raw SQL writes on the writer while writer_unavailable is stale 
A stale RoutingPrismaWrapper.writer_unavailable flag made WriterPinnedClient hand back the routed wrapper, where query_raw is classified as a read, so the access-group UPDATE statements behind /key/regenerate, /key/generate with access_group_ids and model rename/delete went to the read replica and failed with SQLSTATE 25006. Route those raw statements through the underlying writer regardless of the flag; a raw SQL write has no replica fallback. WriterPinnedClient keeps yielding to the replica for degraded reads.

The model sync's backing-row count stays on the writer too: it runs right after the row delete/update on the writer and a lagging replica could still report the removed row, which would leave the group naming a model nobody serves.

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 19:55:07 +00:00
Devin AI
f351685137 fix(proxy): always decrement on spend reset and reseed counters from the DB 
A zero computed decrement still fell back to an absolute spend: 0, so
spend flushed between the read and the commit of a zero-spend row was
erased the same way. The payload is now always
{"spend": {"decrement": spend_decrement}}, and a 0.0 decrement is a
no-op that preserves later spend.

Post-reset the admission spend counter was seeded with the in-memory
post-reset value, which misses increments that raced the reset write.
Invalidate instead: delete the in-memory and Redis counter keys so the
next get_current_spend read reseeds from the committed row.

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 19:52:37 +00:00
yassin
ea6314492f refactor(ui): move key rate limit fields into KeyRateLimitFields 
key_edit_view.tsx crossed the 800 line eslint max-lines ceiling once the
tpd_limit field landed. Move the tpm/rpm/tpd fields into a shared
KeyRateLimitFields control so the edit view stays under the limit

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 19:51:43 +00:00
Yassin Kortam
4d526ad747
Merge pull request #41216 from BerriAI/litellm_cache_custom_hf_tokenizer 
fix(utils): cache custom HuggingFace tokenizers across /utils/token_counter requests
 2026-09-15 12:51:18 -07:00
yassin
c9ccda210e test(proxy): declare org member spend test bindings Final 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 19:49:58 +00:00
Yassin Kortam
d5b96648ea
Merge pull request #41161 from BerriAI/litellm_headroom_protect_cached_prefix 
fix(headroom): protect the cached prefix through the last cache_control breakpoint
 2026-09-15 12:49:50 -07:00
Yassin Kortam
c190f0f4b9
Merge pull request #41125 from BerriAI/litellm_alert_5xx_http_exceptions 
fix(alerting): send llm_exceptions Slack alert for 5xx HTTPException and ProxyException
 2026-09-15 12:48:58 -07:00
Yassin Kortam
a7b894f91b
Merge pull request #41271 from BerriAI/litellm_org_zero_max_budget 
fix(proxy): enforce organization budgets when max_budget is 0
 2026-09-15 12:47:18 -07:00
Devin AI
31052629d6 ci: rerun checks after docs for LITELLM_HTTP2 merged 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 19:46:30 +00:00
Yassin Kortam
60808520df
Merge pull request #40904 from BerriAI/litellm_jwt_agent_id_claim 
feat(proxy): bind JWT claims to registered agents via agent_id_jwt_field
 2026-09-15 12:46:18 -07:00
yassin
ecc2ed0c94 chore: merge main into litellm_lit_1795 and resolve test_router conflict 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 19:45:39 +00:00
Tin Chi Lo
864f4a7a0e feat(auto-router): add per-model Fast mode toggle 2026-09-15 12:45:16 -07:00
yassin
d350aaf8bd Merge remote-tracking branch 'origin/main' into litellm_prometheus_401_failed_requests_metric 2026-09-15 19:43:47 +00:00
Joshua Valluru
eda98f38d9 fix(mcp): preserve canonical JWT owner lookup without cached identity 2026-09-15 12:41:12 -07:00
Yassin Kortam
1ca4579375
Merge pull request #41178 from BerriAI/litellm_request_override_selector_callbacks 
fix(router): bind per-request routing_strategy override selectors to the request's callbacks
 2026-09-15 12:39:20 -07:00
Joshua Valluru
92e182b898 fix(mcp): persist OAuth credentials for validated JWT users 2026-09-15 12:29:03 -07:00
Devin AI
2f33727cc9 fix(proxy): reset budgets by decrementing pre-reset spend instead of zeroing rows 
The budget reset job read a row's spend, reset it in place, then wrote
spend: 0 (or decremented by max_budget under rollover) when committing.
Any spend the batch writer incremented into the row between the read and
the commit was erased while LiteLLM_DailyUserSpend kept it, so the daily
rollup permanently exceeded the counters.

Capture each row's spend before _reset_budget_common mutates it and write
a decrement of pre_spend - post_spend, which equals max_budget in the
rollover-over-cap case it replaces. Rows with no spend still get an
absolute spend: 0.

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 19:24:33 +00:00
Devin AI
93f27d6082 fix(registry): add gemini 3.8 live, azure gpt-5.5/luna snapshots, doubao seed 2.1, fix together v4.1 flash context 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 19:15:46 +00:00
yucheng
d5e056491c fix(guardrails): keep the assistant turn when scoping empties the request history 
A request whose turns all fall outside the guardrail's scope, such as a user-only
request under scan_only_tool_results, still supplied a conversation, so the response
scan now carries the reply as the sole assistant turn instead of dropping
structured_messages. Response-only behavior stays when no conversation was supplied

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 19:12:15 +00:00
Tin Chi Lo
55fc0deabc style(router): format encrypted task selection 2026-09-15 12:09:05 -07:00
Tin Chi Lo
cadb7ee44d fix(router): preserve native encrypted capability tasks 2026-09-15 12:05:30 -07:00
Devin AI
7566164e46 Merge remote-tracking branch 'origin/main' into litellm_registry_audit_2026_09_14 2026-09-15 19:02:41 +00:00
Mateo Wang
d3929287fe
Merge pull request #41189 from BerriAI/litellm_per_turn_control_beta 
fix(anthropic): add the per-turn-control beta when a message carries output_config
 2026-09-15 11:53:01 -07:00
Tin Chi Lo
896f35c751 fix(router): extract capability tasks with request scoped markers 2026-09-15 11:52:20 -07:00
berriai-litellm-provider-info-sync[bot]
9f4fbcfe41
chore(prices): sync Azure prices: 14 models [enrichment failed: Google Gemini, 34 held] 
azure/eu/gpt-5-2025-08-07: input_cost_per_token_batches, input_cost_per_token_priority, output_cost_per_token_batches, output_cost_per_token_priority, cache_read_input_token_cost_priority
azure/eu/gpt-5-mini-2025-08-07: input_cost_per_token_batches, input_cost_per_token_priority, output_cost_per_token_batches, output_cost_per_token_priority, cache_read_input_token_cost_priority
azure/eu/gpt-5-nano-2025-08-07: input_cost_per_token_batches, output_cost_per_token_batches
azure/eu/o1-mini-2024-09-12: 
azure/eu/o1-preview-2024-09-12: 
azure/o1-mini-2024-09-12: input_cost_per_token_batches, output_cost_per_token_batches
azure/us/gpt-4.1-2025-04-14: input_cost_per_token_priority, output_cost_per_token_priority, cache_read_input_token_cost_priority
azure/us/gpt-4.1-mini-2025-04-14: input_cost_per_token_priority, output_cost_per_token_priority, cache_read_input_token_cost_priority
azure/us/gpt-4.1-nano-2025-04-14: cache_read_input_token_cost, input_cost_per_token_batches
azure/us/gpt-5-2025-08-07: input_cost_per_token_batches, input_cost_per_token_priority, output_cost_per_token_batches, output_cost_per_token_priority, cache_read_input_token_cost_priority
azure/us/gpt-5-mini-2025-08-07: input_cost_per_token_batches, input_cost_per_token_priority, output_cost_per_token_batches, output_cost_per_token_priority, cache_read_input_token_cost_priority
azure/us/gpt-5-nano-2025-08-07: input_cost_per_token_batches, output_cost_per_token_batches
azure/us/o1-mini-2024-09-12: 
azure/us/o1-preview-2024-09-12:
 2026-09-15 18:51:21 +00:00
Yassin Kortam
501be3143d fix(proxy): enforce organization budgets when max_budget is 0 
_organization_max_budget_check returned early whenever org_max_budget
was <= 0, so an organization with an explicit max_budget of 0 was
treated as unlimited instead of zero allowance. Key, team, and user
budget checks already skip only on None; align organization budgets
with that convention.

validate_team_org_change had the same defect in a different shape: it
used a truthy check on the org's max_budget when validating a team
move, so an explicit 0 there silently skipped the guard too.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
 2026-09-15 11:49:05 -07:00
berriai-litellm-provider-info-sync[bot]
41caaa301c
chore(prices): sync prices for 2 providers: 235 models, 59 new [enrichment failed: Google Gemini, 34 held] 
azure_ai/Codestral-2501: 
azure_ai/cohere-command-a: 
azure_ai/deepseek-r1: 
azure_ai/deepseek-v3: 
azure_ai/deepseek-v3-0324: 
azure_ai/deepseek-v3.1: 
azure_ai/deepseek-v3.2: 
azure_ai/deepseek-v3.2-speciale: 
azure_ai/deepseek-v4-flash: 
azure_ai/DeepSeek-V4-Flash-0731: 
azure_ai/deepseek-v4-pro: 
azure_ai/embed-v-4-0: 
azure_ai/FW-DeepSeek-V3.2: 
azure_ai/FW-DeepSeek-V4-Pro: 
azure_ai/FW-GLM-5: 
azure_ai/FW-GLM-5.1: 
azure_ai/FW-GLM-5.2: 
azure_ai/FW-GLM-5.2-Fast: input_cost_per_token, output_cost_per_token, cache_read_input_token_cost
azure_ai/FW-Inkling: input_cost_per_token, output_cost_per_token, cache_read_input_token_cost
azure_ai/FW-Kimi-K2.5: 
azure_ai/FW-Kimi-K2.6: 
azure_ai/FW-Kimi-K2.7-Code: 
azure_ai/FW-Kimi-K3: 
azure_ai/FW-MiniMax-M2.5: 
azure_ai/FW-MiniMax-M3: 
azure_ai/FW-Nemotron-3-Ultra-NVFP4: input_cost_per_token, output_cost_per_token, cache_read_input_token_cost
azure_ai/FW-Nemotron-Lightning-3.5-30B-A3B: 
azure_ai/gpt-oss-120b: 
azure_ai/grok-3: 
azure_ai/global/grok-3: 
azure_ai/grok-3-mini: 
azure_ai/global/grok-3-mini: 
azure_ai/grok-4: 
azure_ai/grok-4-1-fast-non-reasoning: 
azure_ai/grok-4-1-fast-reasoning: 
azure_ai/grok-4-20-non-reasoning: 
azure_ai/grok-4-20-reasoning: 
azure_ai/grok-4-fast-non-reasoning: 
azure_ai/grok-4-fast-reasoning: 
azure_ai/grok-4.3: 
azure_ai/grok-4.6: 
azure_ai/grok-code-fast-1: 
azure_ai/kimi-k2.5: 
azure_ai/kimi-k2.6: 
azure_ai/kimi-k2.7-code: 
azure_ai/Llama-3.3-70B-Instruct: 
azure_ai/Llama-4-Maverick-17B-128E-Instruct-FP8: input_cost_per_token, output_cost_per_token
azure_ai/MAI-DS-R1: 
azure_ai/MAI-Image-2.5: 
azure_ai/MAI-Image-2.5-Flash: 
azure_ai/MAI-Image-2e: 
azure_ai/MAI-Thinking-1: 
azure_ai/mistral-large-3: 
azure_ai/Phi-3-medium-128k-instruct: 
azure_ai/Phi-3-medium-4k-instruct: 
azure_ai/Phi-3-mini-128k-instruct: 
azure_ai/Phi-3-mini-4k-instruct: 
azure_ai/Phi-3-small-128k-instruct: 
azure_ai/Phi-3-small-8k-instruct: 
azure_ai/Phi-3.5-mini-instruct:
 2026-09-15 18:46:36 +00:00
Tin Chi Lo
e62f0d0376 fix(router): reject unknown capability policy fields 2026-09-15 11:43:04 -07:00
berriai-litellm-provider-info-sync[bot]
a511c9d45d
chore(prices): sync Google Gemini prices: 7 models [enrichment failed: Google Gemini, 86 held] 
gemini/gemini-3.1-flash-image: 
gemini/gemini-3.1-flash-lite: cache_read_input_audio_token_cost, input_cost_per_audio_token_batches
gemini/gemini-3.1-flash-lite-image: 
gemini-3.1-flash-live-preview: 
gemini/gemini-3.1-flash-live-preview: 
gemini/gemini-3.1-flash-tts-preview: output_cost_per_audio_token, input_cost_per_token_batches
gemini/gemini-3.1-pro-preview: input_cost_per_token_flex, output_cost_per_token_flex, cache_read_input_token_cost_flex
 2026-09-15 18:41:29 +00:00
mateo-berri
f490172338 test(anthropic): drop docstrings and wrap a long line in the per-turn-control tests 2026-09-15 11:39:01 -07:00
Mateo Wang
2e06d195b2
Merge pull request #39857 from BerriAI/litellm_e2e_reliability_module_cells 
test(e2e): cover the reliability retry, cooldown, fallback, and routing-strategy cells
 2026-09-15 11:38:51 -07:00
Tin Chi Lo
914ae9b248 test(ui): use the current deployment affinity label 2026-09-15 11:38:28 -07:00
yassin
fb00567e4c fix(proxy): track per-member organization spend so the Organizations UI shows member spend 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 18:38:07 +00:00
berriai-litellm-provider-info-sync[bot]
1517f1205c
chore(prices): sync Google Gemini prices: 10 models [enrichment failed: Google Gemini, 177 held] 
gemini/gemini-2.5-computer-use-preview-10-2025: 
gemini/gemini-2.5-flash: input_cost_per_token_flex, output_cost_per_token_flex, input_cost_per_token_batches, input_cost_per_token_priority, output_cost_per_token_batches, output_cost_per_token_priority, cache_read_input_token_cost_flex, cache_read_input_audio_token_cost, input_cost_per_audio_token_batches, cache_read_input_token_cost_priority
gemini/gemini-2.5-flash-image: input_cost_per_token_flex, input_cost_per_token_batches, input_cost_per_token_priority
gemini/gemini-2.5-flash-lite: input_cost_per_token_flex, output_cost_per_token_flex, input_cost_per_token_batches, input_cost_per_token_priority, output_cost_per_token_batches, output_cost_per_token_priority, cache_read_input_token_cost_flex, cache_read_input_audio_token_cost, input_cost_per_audio_token_batches, cache_read_input_token_cost_priority
gemini-2.5-flash-preview-tts: 
gemini/gemini-2.5-flash-preview-tts: 
gemini/gemini-2.5-pro: input_cost_per_token_flex, output_cost_per_token_flex, input_cost_per_token_batches, input_cost_per_token_priority, output_cost_per_token_batches, output_cost_per_token_priority, cache_read_input_token_cost_flex, cache_read_input_token_cost_priority, input_cost_per_token_above_200k_tokens_priority, output_cost_per_token_above_200k_tokens_priority, cache_read_input_token_cost_above_200k_tokens_priority
gemini/gemini-2.5-pro-preview-tts: output_cost_per_audio_token, input_cost_per_token_batches
gemini/gemini-3-flash-preview: input_cost_per_token_flex, output_cost_per_token_flex, input_cost_per_token_batches, output_cost_per_token_batches, cache_read_input_token_cost_flex, cache_read_input_audio_token_cost, input_cost_per_audio_token_batches
gemini/gemini-3-pro-image: input_cost_per_token_flex, output_cost_per_token_flex, input_cost_per_token_priority, output_cost_per_token_priority
 2026-09-15 18:36:52 +00:00
Yassin Kortam
41b5d47c71
Merge pull request #41144 from BerriAI/litellm_responses_bridge_filters_unknown_params 
fix(responses): filter bridged kwargs like the native Responses path
 2026-09-15 11:32:38 -07:00
Devin AI
8978b4562f test(main): drop unrelated reformatting 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 18:26:54 +00:00
Yassin Kortam
837423237a
Merge pull request #36775 from MvdB/litellm_presidio_new_entities 
feat(guardrails): add new upstream presidio pii entities including german set
 2026-09-15 11:26:21 -07:00
yujonglee
33000d7e25
Merge pull request #41180 from BerriAI/litellm_rust_bridge_native_stub 
build(rust-bridge): add typed _native stub and validate it with mypy.stubtest
 2026-09-15 11:25:56 -07:00
Devin AI
fd2fb4c44e fix(http): address review on outbound HTTP/2 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 18:25:26 +00:00
Devin AI
da7853c20a test: drop tests that pin vendor facts and add the CLAUDE.md rule 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-09-15 18:24:30 +00:00
Tin Chi Lo
2da9bbfc0f chore: merge main into capability classifier copy 2026-09-15 11:24:01 -07:00
tin-berri
3ac79757f4
Merge pull request #41175 from BerriAI/litellm_team_member_auto_routers 
feat(auto-router): allow opted-in team members to manage their routers
 2026-09-15 11:19:58 -07:00