Model Library

Build with leading open-source models

Explore top-performing models across chat, vision, code, and audio - optimized for production deployment

Start building nowRequest dedicated endpoint

Featured

models

Chat

DeepSeek-R1-0528

Upgraded DeepSeek-R1 with better reasoning, function calling, and coding, using 23K-token thinking to score 87.5% on AIME.

TRY THIS MODEL

Chat

Kimi K2 Instruct-0905

Upgraded state-of-the-art mixture-of-experts agentic intelligence model with 1T parameters, 256K context, and native tool use

TRY THIS MODEL

Chat

Qwen3 235B A22B Instruct 2507 FP8

235B MoE model with 22B activation featuring enhanced instruction following, reasoning, and 262K context for cost-efficient high-throughput inference.

TRY THIS MODEL

Chat

Llama 4 Maverick

SOTA 128-expert MoE powerhouse for multilingual image/text understanding, creative writing, and enterprise-scale applications.

TRY THIS MODEL

Audio

Arcana v2

Expressive TTS with 300+ voices across English, Spanish, French, German with multi-lingual code-switching and paralinguistic features.

TRY THIS MODEL

new

Chat

Gemma 3 27B

Lightweight model with vision-language input, multilingual support, visual reasoning, and top-tier performance per size.

TRY THIS MODEL

Chat

DeepSeek-V3-0324

DeepSeek's latest open Mixture-of-Experts model challenging top AI models at much lower cost.

TRY THIS MODEL

Chat

DeepSeek-V3.1

671B parameters (37B activated), 128K context, hybrid thinking/non-thinking modes, advanced tool calling, agent capabilities

TRY THIS MODEL

All

Chat

MiniMax M2.5

SOTA coding and agentic model achieving 80.2% SWE-Bench Verified, trained on 200K+ real-world environments with architect-level planning, full-stack development, and office deliverables at 10% the cost of frontier models.

TRY THIS MODEL

new
minimaxai

Reasoning

Code

Code

Qwen3.5-397B-A17B

Native multimodal model with 397B parameters (17B activated), 256K context, achieving 87.8% MMLU-Pro and 201-language support with hybrid Gated Delta Networks architecture.

TRY THIS MODEL

new
Qwen

Code

Chat

Image

Nano Banana Pro (Gemini 3 Pro Image)

Gemini 3 Pro Image model with SOTA text rendering, dramatically cleaner details, 14-image composition, precise editing controls for infographics and layouts at 2K/4K resolution

TRY THIS MODEL

New
Google

Image

Chat

Kimi K2.5

1T parameter native multimodal thinking agent achieving 50.2% HLE with tools, Agent Swarm orchestration, vision-grounded coding, 256K context, and native INT4 quantization for 2x speed.

TRY THIS MODEL

New
Moonshot AI

Chat

Code

LLM

Chat

GLM-5

Open-source foundation model for complex systems engineering and long-horizon agent workflows, with Opus-level intelligence for multi-stage programming tasks and deep debugging.

TRY THIS MODEL

new
ZAI

Chat

Reasoning

LLM

Chat

gpt-oss-120B

120B parameters, 128K context, reasoning with chain-of-thought, MoE architecture, Apache 2.0 license

TRY THIS MODEL

OpenAI

Reasoning

Code

DeepSeek-V3.2-Exp

685B parameter experimental model with 128K context, DeepSeek Sparse Attention for long-context efficiency, performance parity with V3.1-Terminus

TRY THIS MODEL

DeepSeek

Code

LLM

Reasoning

Chat

Ministral 3 8B Instruct 2512

Efficient 8B multimodal model with 256K context, balanced quality and cost for everyday assistants, multimodal copilots, and mid-scale production agents.

TRY THIS MODEL

new
Mistral AI

Chat

Audio

MiniMax Speech 2.6 Turbo

Ultra-low latency TTS model with under 250ms end-to-end latency, seamless format handling for URLs, emails, dates, and Fluent voice cloning across 40+ languages.

TRY THIS MODEL

new
minimaxai

Audio

Code

Qwen3-Coder-Next

80B parameter (3B activated) coding agent achieving 74.2% SWE-Bench Verified, 256K context, and advanced tool calling for robust agentic workflows.

TRY THIS MODEL

new
Qwen

Code

Chat

LLM

Chat

Gemma 3 27B

Lightweight model with vision-language input, multilingual support, visual reasoning, and top-tier performance per size.

TRY THIS MODEL

Google

Chat

Code

Vision

Chat

Llama 4 Maverick

SOTA 128-expert MoE powerhouse for multilingual image/text understanding, creative writing, and enterprise-scale applications.

TRY THIS MODEL

Meta

Chat

Vision

Chat

Qwen3 235B A22B Instruct 2507 FP8

235B MoE model with 22B activation featuring enhanced instruction following, reasoning, and 262K context for cost-efficient high-throughput inference.

TRY THIS MODEL

Qwen

Chat

Reasoning

Video

Google Veo 3.0

Latest Veo generation with improved quality, 720p resolution, 8-second duration.

TRY THIS MODEL

Google

Video

Image

FLUX.2 [pro]

Production-grade image model supporting up to 4MP outputs, 8 reference images on API, 10 on Playground, with photorealistic detail, hex color accuracy, multi-reference character consistency

TRY THIS MODEL

Black Forest Labs

Image

Chat

Kimi K2 Instruct-0905

Upgraded state-of-the-art mixture-of-experts agentic intelligence model with 1T parameters, 256K context, and native tool use

TRY THIS MODEL

Moonshot AI

LLM

Video

Sora 2 Pro

Professional-grade video AI with extended 16-second generation, superior quality, no watermarks, and enhanced physics simulation for commercial use.

TRY THIS MODEL

OpenAI

Video

Chat

GLM-4.7

200K context reasoning model ranking #1 open-source on LMArena Code Arena, with 73.8% SWE-bench Verified, enhanced agentic coding, and advanced interleaved thinking for stable multi-step execution.

TRY THIS MODEL

ZAI

Chat

Reasoning

LLM

Chat

DeepSeek-V3.1

671B parameters (37B activated), 128K context, hybrid thinking/non-thinking modes, advanced tool calling, agent capabilities

TRY THIS MODEL

DeepSeek

LLM

Chat

Reasoning

Audio

Arcana V3

Multilingual TTS with native code-switching across 11 languages, preserving prosody and accent consistency for natural transitions in global voice applications.

TRY THIS MODEL

new
Rime

Audio

No search results.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Run any model on the fastest endpoints

Use our API to deploy any open-source model on the fastest inference stack available with optimal cost efficiency.

Scale into a dedicated deployment anytime with a custom number of instances to get optimal throughput.

Start buildingRequest dedicated endpoint
Curl
Python
Typescript

RUN INFERENCE

curl -X POST "https://api.together.xyz/v1/chat/completions" \
  -H "Authorization: Bearer $TOGETHER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
	"model": "meta-llama/Llama-Vision-Free",
	"messages": [{"role": "user", "content": "What are some fun things to do in New York?"}]
  }'

RUN INFERENCE

from together import Together

client = Together()

response = client.chat.completions.create(
    model="meta-llama/Llama-Vision-Free",
    messages=[{"role": "user", "content": "What are some fun things to do in New York?"}],
)

print(response.choices[0].message.content)

RUN INFERENCE

import Together from "together-ai";

const together = new Together({ apiKey: process.env.TOGETHER_API_KEY });

const response = await together.chat.completions.create({
    messages: [{"role": "user", "content": "What are some fun things to do in New York?"}],
    model: "meta-llama/Llama-Vision-Free",
});

console.log(response.choices[0].message.content)

Subscribe to newsletter

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.