fix(deepseek): correct max_output_tokens for deepseek-v4 models - #31018

Open

New1Direction wants to merge 21 commits into

BerriAI:litellm_oss_stagingfrom

New1Direction:fix/deepseek-v4-max-output-tokens

Commits on Jun 22, 2026

fix: correct amazon.titan-embed-text-v2 input price to $0.02/1M tokens (#29693)
Srivatsa03
authored
343ef87
fix(sensitive_data_masker): fully mask secrets at or below the reveal threshold (#30764)
Srivatsa03
authored
45b4dca
refactor(fireworks_ai): remove deprecated audio transcriptions endpoint (#30917)
shzdehmd
authored
2b29b92
feat: add darkbloom provider (#30876)
jychp
authored
19450bd
fix: fail fast for non-Postgres database URLs (#30883)
krishvsoni
authored
488f787
fix(bedrock): surface modeled HTTP status for mid-stream error events so 5xx is retryable (#24608) (#30946)
kingdoooo
authored
67c0183
fix(mcp): re-land native tool preservation with typed annotations (#30645)
ayushh0110
authored
0a359ce
fix(sambanova): return embeddings supported params instead of dropping them (#30937)
hoangsonww
authored
e5ce7ab
fix(router): send fallback metadata when streaming (#30914)
talmar-island
authored
7fa04db
fix(mistral): drop output-only reasoning fields from input messages (#30884)

hemant1026
and
cursoragent
authored
37a2a32
fix(perplexity): bill search queries at the per-request price, not 1/1000 of it (#30652)
devYRPauli
authored
8f4a2bc
fix: include model_access_groups when expanding all-team-models in get_team_models (#30622)
factnn
authored
ad2b395
fix(fireworks_ai): return None for transcription in get_supported_openai_params
Sameerlite
committed
da85887
fix(proxy): gate include_fallback_errors behind expose_fallback_errors_to_caller setting
Sameerlite
committed
e7ff3e1
test(proxy): opt in to expose_fallback_errors_to_caller in streaming SSE test
Sameerlite
committed
db2bc49
test(prompt_templates): make test_convert_url hermetic instead of hitting picsum.photos
mateo-berri
committed
0a82e13
fix(interactions): drop role from Interaction response to match Google spec
mateo-berri
committed
5663bcf
fix: align all-team-models sentinel access
cursoragent
committed
d093dc2
fix(router): forward include_fallback_errors through multi-hop fallbacks
mateo-berri
committed
dc6611b
fix(deepseek): correct max_output_tokens for deepseek-v4 models
New1Direction
committed
11d6b34

Commits on Jun 23, 2026

test(deepseek): assert max_output_tokens for deepseek-v4 models
New1Direction
committed
33086d9

Navigation Menu

Uh oh!

fix(deepseek): correct max_output_tokens for deepseek-v4 models - #31018

fix(deepseek): correct max_output_tokens for deepseek-v4 models#31018
New1Direction wants to merge 21 commits into
BerriAI:litellm_oss_stagingfrom
New1Direction:fix/deepseek-v4-max-output-tokens

Commits on Jun 22, 2026

fix: correct amazon.titan-embed-text-v2 input price to $0.02/1M tokens (#29693)

fix(sensitive_data_masker): fully mask secrets at or below the reveal threshold (#30764)

refactor(fireworks_ai): remove deprecated audio transcriptions endpoint (#30917)

feat: add darkbloom provider (#30876)

fix: fail fast for non-Postgres database URLs (#30883)

fix(bedrock): surface modeled HTTP status for mid-stream error events so 5xx is retryable (#24608) (#30946)

fix(mcp): re-land native tool preservation with typed annotations (#30645)

fix(sambanova): return embeddings supported params instead of dropping them (#30937)

fix(router): send fallback metadata when streaming (#30914)

fix(mistral): drop output-only reasoning fields from input messages (#30884)

fix(perplexity): bill search queries at the per-request price, not 1/1000 of it (#30652)

fix: include model_access_groups when expanding all-team-models in get_team_models (#30622)

fix(fireworks_ai): return None for transcription in get_supported_openai_params

fix(proxy): gate include_fallback_errors behind expose_fallback_errors_to_caller setting

test(proxy): opt in to expose_fallback_errors_to_caller in streaming SSE test

test(prompt_templates): make test_convert_url hermetic instead of hitting picsum.photos

fix(interactions): drop role from Interaction response to match Google spec

fix: align all-team-models sentinel access

fix(router): forward include_fallback_errors through multi-hop fallbacks

fix(deepseek): correct max_output_tokens for deepseek-v4 models

Commits on Jun 23, 2026

test(deepseek): assert max_output_tokens for deepseek-v4 models