scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-10-07 02:59:05 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 103
46522 commits 3222 branches 1436 tags 4.7 GiB
Commit graph

46522 commits

This branch
This branch
All branches
Author SHA1 Message Date
mateo-berri
de1f38820a fix(passthrough): flush interrupted streams on client disconnect and reuse cached gigachat http clients 2026-08-30 13:36:51 -07:00
mateo-berri
14f392bb9b fix(docker): bump wolfi-base for glibc 2.44 and pin apk python to 3.13 2026-08-30 13:11:32 -07:00
mateo-berri
a5fa8ebfa7 fix(passthrough): keep upstream error body readable for streaming error status mapping 2026-08-30 12:59:18 -07:00
mateo-berri
db1e0717f9 fix(guardrail_translation): assemble responses stream text from delta events for terminal-failure scans 2026-08-30 12:52:03 -07:00
mateo-berri
fb9ec79d7c Merge branch 'litellm_internal_staging' into litellm_decrease_anys_opus5_r2 2026-08-30 12:49:29 -07:00
mateo-berri
99a6dd02af fix(proxy): narrow audio_speech response before reading upstream content-type 2026-08-30 12:46:11 -07:00
mateo-berri
1f702f50ad Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_fix_openai_embedding_encoding_format_omit 
Staging moved again mid-recovery; only the ANN201 ratchet conflicted and this branch's tighter limit stands.
 2026-08-30 12:43:40 -07:00
mateo-berri
24c5846c75 Merge branch 'litellm_internal_staging' into litellm_fix_bedrock_buffered_responses_stream 2026-08-30 12:35:39 -07:00
mateo-berri
611750cd11 Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_fix_master_key_rotation_blocked 
# Conflicts:
#	litellm/proxy/management_endpoints/key_management_endpoints.py
 2026-08-30 12:35:21 -07:00
mateo-berri
11e0502239 Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_fix_openai_embedding_encoding_format_omit 
Resolves budget-ratchet conflicts by taking staging's tighter limits and reworks the embedding raw-response helpers so the branch stays net-negative on the LIT001/LIT002 ceilings staging lowered: the request methods now return the LegacyAPIResponse and each caller keeps a single dict(headers) conversion.
 2026-08-30 12:32:49 -07:00
mateo-berri
f04bfa457a Merge branch 'litellm_internal_staging' into litellm_fix_chat_anyof_tool_schema 2026-08-30 12:31:11 -07:00
mateo-berri
ce52e39052 fix(gigachat): honor ssl_verify config on router passthrough and type the request body 2026-08-30 12:28:41 -07:00
mateo-berri
739f61df7d test(model_management): drop docstring that restates the serialization path 2026-08-30 12:26:06 -07:00
mateo-berri
c236bcf241 Merge branch 'litellm_internal_staging' into litellm_decrease_anys_opus5_r2 2026-08-30 12:16:13 -07:00
Mateo Wang
4ba8517134
Merge pull request #38722 from BerriAI/litellm_bedrock_guardrail_stream_audit 
feat(bedrock): honor streaming buffer/sampling config for unbuffered post_call scans
 2026-08-30 10:17:11 -07:00
Mateo Wang
8a156ed42d
Merge pull request #36722 from BerriAI/litellm_decrease_anys_fable6 
chore(typing): clear 1.2k basedpyright Any errors across 16 hotspot files
 2026-08-30 10:09:02 -07:00
mateo-berri
2420e3f202 Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_fix_gemini_tts_container 2026-08-30 10:02:43 -07:00
mateo-berri
fbf7644676 Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_decrease_anys_fable6 
# Conflicts:
#	basedpyright-code-budget.json
#	litellm/integrations/websearch_interception/handler.py
#	litellm/proxy/response_polling/background_streaming.py
#	ruff-strict-budget.json
#	type-discipline-budget.json
 2026-08-30 10:02:06 -07:00
mateo-berri
b6bd749c02 fix(ui): withhold team-scoped model writes from view-only sessions too 
The route-level RBAC in litellm/proxy/auth/route_checks.py 403s
/model/new, /model/update, and /model/delete for proxy_admin_viewer on
the session role alone, before ModelManagementAuthChecks' team-admin
carve-out can run. A view-only session therefore gets no model write
affordance, team admin or not.
 2026-08-30 10:01:44 -07:00
Devin AI
5c7e6b80c9 test: isolate global MCP registry and pin savings tests to bundled cost map 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-08-30 10:18:15 +00:00
Devin AI
693279afb4 chore(techdebt): clear fresh debt from the 2026-08-29 window 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-08-30 07:57:16 +00:00
Devin AI
36c53e1288 chore(ui): regenerate schema.d.ts for updated endpoint descriptions 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-08-30 07:46:45 +00:00
Devin AI
4fb1440747 docs(proxy): clarify spend semantics on /v2/user/info and /user/daily/activity 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
 2026-08-30 07:40:56 +00:00
siyoon
3822ecc8ff chore(tests): drop redundant capability comment 
Per greptile review + CLAUDE.md comment policy: the comment restated
the immediately following assertion without adding value.
 2026-08-30 14:29:42 +09:00
siyoon
e9f1af8473 chore(tests): drop redundant capability comment 
Per greptile review + CLAUDE.md comment policy: the comment restated
the immediately following assertions without adding value.
 2026-08-30 14:29:28 +09:00
siyoon
e7bfe99cd3 feat(friendli): add zai-org/GLM-5.3 model pricing 
Per https://api.friendli.ai/serverless/v1/models:
- $1.40 input / $4.40 output / $0.26 cached input per MTok
- 1M context, 1M max output, reasoning with effort low/high/max
  (per HF chat_template.jinja: low/high honored, anything else -> max)
- tool calling, parallel tool calls, structured output, prompt caching
- text-only (no vision), flagship GLM model
 2026-08-30 14:23:36 +09:00
siyoon
3ea4b715ba feat(friendli): add zai-org/GLM-5.3-Flash model pricing 
Per https://api.friendli.ai/serverless/v1/models:
- $0.15 input / $0.50 output / $0.03 cached input per MTok
- 1M context, 1M max output, reasoning with effort low/high/max
  (per HF chat_template.jinja: low/high honored, anything else -> max)
- tool calling, parallel tool calls, structured output, prompt caching
- image + video input (native multimodal)
 2026-08-30 14:22:41 +09:00
mateo-berri
a928c1429e fix(proxy): preserve model table columns on master key rotation 2026-08-29 22:12:35 -07:00
mateo-berri
eac5dc10f3 fix(guardrails): apply PUT /guardrails/{id} to the serving worker immediately and reject invalid configs with 422 2026-08-29 22:11:36 -07:00
mateo-berri
b0ce17c755 fix(gigachat): generic env-credential passthrough fallback plus type hardening 
- forward unrouted /gigachat/* requests with env credentials like other passthrough providers (the old fallback returned 400 on any request without a routed model, /gigachat/models included)
- fix basedpyright budget breaches across the gigachat provider, common_request_processing, and llm_passthrough_endpoints with real narrowing, no new suppressions
- add regression tests for the fallback target, auth header, and model-less endpoints
 2026-08-29 22:08:54 -07:00
mateo-berri
70e2f4e68f Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_gigachat_passthrough_25886 
# Conflicts:
#	litellm/llms/gigachat/chat/transformation.py
 2026-08-29 22:08:54 -07:00
mateo-berri
60296cb540 Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_bedrock_guardrail_stream_audit 2026-08-29 21:49:13 -07:00
Mateo Wang
5e4b3838aa
Merge pull request #37778 from BerriAI/litellm_decrease_anys_opus5 
chore(typing): clear Any seams across 47 files, ratchet basedpyright ceilings -3,302
 2026-08-29 21:48:11 -07:00
mateo-berri
667f761f1d Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_decrease_anys_fable6 2026-08-29 21:43:56 -07:00
mateo-berri
43c838f4b9 Merge branch 'litellm_internal_staging' into litellm_decrease_anys_opus5 2026-08-29 21:41:04 -07:00
mateo-berri
74fb398f9b fix(ui): hide model write affordances from view-only admin sessions 2026-08-29 21:40:28 -07:00
Mateo Wang
b63bced31b
Merge pull request #38863 from BerriAI/litellm_fix_savings_test_cache_read_premise 
fix(tests): derive the no-cache-read-rate savings baseline from the model map
 2026-08-29 21:40:18 -07:00
mateo-berri
fd72ae830c test(model_management): drive /model/block and /model/unblock through response serialization 
The route-level regression test returns a real prisma row from a mocked
update and asserts both routes serialize it to a 200 with the toggled
blocked flag, which is exactly the path that raised AttributeError before
the validator guard. Also binds the loop variable in the e2e poll lambda
(ruff B023).
 2026-08-29 21:39:07 -07:00
mateo-berri
22ff5a8a69 Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_decrease_anys_fable6 
# Conflicts:
#	basedpyright-code-budget.json
#	ruff-strict-budget.json
#	type-discipline-budget.json
 2026-08-29 21:32:59 -07:00
mateo-berri
b418ccd738 fix(azure): flatten top-level tool schema combinators on Azure chat completions 
Azure's chat completions validator rejects tool parameters carrying a
top-level anyOf/oneOf/allOf for every model family. AzureOpenAIConfig and
the o-series config now flatten them via the shared helper moved to
prompt_templates common_utils. Requests bridged to the Responses API for
gpt-5.4+ with reasoning active keep the union, which that surface accepts
 2026-08-29 21:27:57 -07:00
mateo-berri
a23f0fc3c3 Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_bedrock_guardrail_stream_audit 
# Conflicts:
#	litellm/proxy/common_request_processing.py
 2026-08-29 21:23:35 -07:00
samzong
26e71ddc54 fix(proxy): serialize model block responses 
Signed-off-by: samzong <samzong.lu@gmail.com>
 2026-08-29 21:22:17 -07:00
mateo-berri
35a375e26f fix(speech): stop vertex gemini tts from dropping response_format in cloud tts param mapping 2026-08-29 21:10:15 -07:00
mateo-berri
b67b44bdaa fix(proxy): map audio_speech errors to their status codes instead of a blanket 500 2026-08-29 21:10:14 -07:00
mateo-berri
0e27e09fae Merge branch 'litellm_internal_staging' into litellm_fix_chat_anyof_tool_schema 2026-08-29 20:57:35 -07:00
Tin Chi Lo
f62aa1b3a8 fix(tests): derive the no-cache-read-rate savings baseline from the model map 2026-08-29 19:35:10 -07:00
yucheng-berri
d44d281d1d
fix(proxy): emit timing headers and overhead for /v1/messages and /v1/responses (#38840) 2026-08-29 18:11:58 -07:00
yuneng-jiang
10631eb834
Merge pull request #38805 from BerriAI/litellm_internal_staging 
chore(ci): promote internal staging to main
 2026-08-29 18:09:58 -07:00
yuneng-jiang
6b33d17563
Merge pull request #38850 from BerriAI/litellm_e2e_retry_transient_upstream 
test(e2e): retry upstream-saturation failures in the claude CLI driver
 2026-08-29 17:48:04 -07:00
yuneng-jiang
df848d85ff
Merge pull request #38833 from BerriAI/litellm_deflake_reliability_fallbacks 
test(e2e): stop the reliability fallback tests flaking on gpt-5.5's reasoning budget
 2026-08-29 17:44:30 -07:00