Together AI Python SDK v2.0 Release Candidate — type-safe client
AI Impact Summary
Together AI is releasing Python SDK v2.0, a new, type-safe client designed to replace the legacy v1.0 package. This release introduces significant improvements including a redesigned error model with specific HTTP status code errors, new Jobs and Hardware APIs, and enhanced debugging tools like .with_raw_response and .with_streaming_response helpers. The migration is recommended to take advantage of the new features and improved stability, and the release includes beta APIs for Instant Clusters.
Together AI Python SDK v2.0 Release Candidate — type-safe client
Organizations using the legacy Together AI Python SDK v1.0 must migrate to v2.0 to access new features, improved error handling, and optimized performance, ensuring continued compatibility with the platform's evolving capabilities.
Models affected
new
Together Python SDK v2.0
sdk
updated
OpenAPI
api
updated
Chat Completions API
api
updated
Embeddings API
api
updated
Assistants API
tool
updated
Python
sdk
deprecated
Qwen/Qwen3-Coder-Next-FP8
model
deprecated
deepseek-ai/DeepSeek-R1
model
deprecated
Qwen/Qwen3-235B-A22B-fp8-tput
model
deprecated
meta-llama/Llama-4-Scout-17B-16E-Instruct
model
deprecated
meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo
model
deprecated
BAAI/bge-large-en-v1.5
model
deprecated
deepcogito/cogito-v2-1-671b
model
deprecated
google/gemma-3n-E4B-it
model
deprecated
google/gemma-4-31B-it
model
deprecated
google/gemma-4-26B-A4B-it
model
deprecated
google/veo-3.1-test-debug
model
deprecated
vidu/vidu-q3
model
deprecated
vidu/vidu-q3-turbo
model
deprecated
Wan-AI/wan2.7-i2v
model
deprecated
Wan-AI/wan2.7-r2v
model
deprecated
Qwen/Qwen3-VL-8B-Instruct
model
deprecated
Qwen/Qwen3-235B-A22B-Thinking-2507
model
deprecated
Qwen/Qwen2.5-72B-Instruct-Turbo
model
deprecated
meta-llama/Llama-Guard-3-11B-Vision-Turbo
model
deprecated
meta-llama/LlamaGuard-2-8b
model
deprecated
marin-community/marin-8b-instruct
model
deprecated
nvidia/NVIDIA-Nemotron-Nano-9B-v2
model
Risk domains
795%780%
Source text
Together AI is releasing the
Python SDK v2.0 Release Candidate
— a new, OpenAPI‑generated, strongly‑typed client that replaces the legacy v1.0 package and brings the SDK into lock‑step with the latest platform features.
pip install together==2.0.0a9
RC Period:
The v2.0 RC window starts today and will run for
approximately 1 month
. During this time we’ll iterate quickly based on developer feedback and may make a few small, well‑documented breaking changes before GA.
Type‑Safe, Modern Client:
Stronger typing across parameters and responses, keyword‑only arguments, explicit
NOT_GIVEN
handling for optional fields, and rich
together.types.*
definitions for chat messages, eval parameters, and more.
Redesigned Error Model:
Replaces
TogetherException
with a new
TogetherError
hierarchy, including
APIStatusError
and specific HTTP status code errors such as
BadRequestError (400)
,
AuthenticationError (401)
,
RateLimitError (429)
, and
InternalServerError (5xx)
, plus transport (
APIConnectionError
,
APITimeoutError
) and validation (
APIResponseValidationError
) errors.
New Jobs API:
Adds first‑class support for the
Jobs API
(
client.jobs.*
) so you can create, list, and inspect asynchronous jobs directly from the SDK without custom HTTP wrappers.
New Hardware API:
Adds the
Hardware API
(
client.hardware.*
) to discover available hardware, filter by model compatibility, and compute effective hourly pricing from
cents_per_minute
.
Raw Response & Streaming Helpers:
New
.with_raw_response
and
.with_streaming_response
helpers make it easier to debug, inspect headers and status codes, and stream completions via context managers with automatic cleanup.
Code Interpreter Sessions:
Adds session management for the
Code Interpreter
(
client.code_interpreter.sessions.*
), enabling multi‑step, stateful code‑execution workflows that were not possible in the legacy SDK.
High Compatibility for Core APIs:
Most core usage patterns, including
chat.completions
,
completions
,
embeddings
,
images