Pending Signals

Web search is available on the Claude API for $10 per 1,000 searches , plus...

Claude Haiku 4.5

Claude Opus 4.6

1 - See the pricing page for complete pricing information including batch...

max_tokens

Claude Sonnet 4.6

Feb 23, 2026 Announcements Detecting and preventing distillation attacks

Mar 11, 2026 Announcements Introducing The Anthropic Institute

Same as Standard, Batch pricing not yet implemented $0.05 (text / image /...

text-embeddings-004

Released gemini-2.5-pro-exp-03-25 , a public experimental Gemini model with...

The following Gemini Live API models are now shut down:...

allow_adult setting in Image to Video generation are now available in...

Introduced increased rate limits for the upcoming paid tier of Gemini 1.5 Pro.

GenerateContent is a single unified endpoint for chat and text.

Released Gemini 2.0 Flash Experimental for public preview. Gemini 2.0 Flash...

total_thought_tokens

General availability release for Gemini 1.5 Pro's 2M context window.

Gemini 1.0 Pro Vision deprecated.

Updated candidate count: Gemini models only return 1 candidate.

gemini-2.5-flash-preview-09-2025

Gemini 2.5 Pro TTS Preview

Gemini 2.0 Flash-Lite Deprecated

Our second generation workhorse model, with next-gen features and improved...

Gemini models are available in either stable , preview , latest , or...

Points to a preview model which may be used for production. Preview models...

Points to an experimental model which will typically be not be suitable for...

Experimental models are not stable and availability of model endpoints is...

Let’s stay in touch. Get the latest news from Google in your inbox.

Introducing Gemini 3

Model Customization

Mistral Large 3

Mistral Medium 3.1

Test the endpoints live

{ "choices" : [ { "finish_reason" : "stop" , "index" : "0" , "message" : { }...

max_tokens

Deprecated Fine Tuning

MCP tools The Agents API SDK can also leverage tools built on the Model...

Agent handoffs Once agents are created, define which agents can hand off...

Web search The Agents API offers web search as a connector, enabling...

Direct Access: Start a conversation by directly specifying the model and...

command-a-vision-07-2025

Command A is available today on the Cohere Platform , HuggingFace , or...

Unified Embeddings produced from mixed modality input (i.e. a single payload...

API Endpoint : Chat API

For production deployment without rate limits, provision a dedicated Model...

Deprecated Features and API Endpoints:

Maximum Output : 8K tokens

Where do I find pricing for our legacy models? For existing customers:...

Learn how we can get your AI into production — quickly and confidently

Command A is our most efficient and performant model to date, specializing...

Understand how Command works with Embed and Rerank to surface verifiable...

curl https://api.groq.com/openai/v1/chat/completions \ -H "Authorization:...

curl "https://api.groq.com/openai/v1/chat/completions" \ -X POST \ -H...

Remote Model Context Protocol (MCP) server integration is now available in...

Universal interface - Connect to thousands of remote MCP servers and tools...

Compound ( groq/compound ) and Compound Mini ( groq/compound-mini ) are...

curl "https://api.groq.com/openai/v1/chat/completions" \ -X POST \ -H...

Added optional chat completion metadata parameter for better compatibility...

Llama 4 Scout ( meta-llama/llama-4-scout-17b-16e-instruct ): Currently 607...

Added country field to the search_settings parameter for agentic tool...

Higher effective rate limits - cached tokens don't count toward rate limits

Production-grade stability - General availability with increased rate limits...

Updated chat completion model parameter to remove deprecated models and add...

Query: "What's the weather in NYC and LA?" Call 1:...

parameters

The updated messages array is then sent back to the model for the next step

With traditional inference speeds of 10-30 tokens/second, multi-tool...

openai/gpt-oss-safeguard-20b

import OpenAI from "openai" ; const client = new OpenAI ({ apiKey:...

Custom functions you define that the model can invoke

Grok is an AI modeled after the Hitchhiker’s Guide to the Galaxy. It is...

Expanding ‘xAI For Government’ with more accessible AI tools for the Federal...

Grok-1.5 comes with improved reasoning capabilities and a context length of...

We are releasing the weights and architecture of our 314 billion parameter...

For more detailed instructions, please refer to the guide Use Context Caching .

DeepSeek API’s Concurrency and Rate Limits

Improve model output performance through Few-shot learning.

prompt_cache_miss_tokens: Number of tokens from the input that were not...

Hint 1: The API price has been updated. For details, please refer to Models...

The disk caching service is now available for all users, requiring no code...

Two new fields in the API response's usage section help users monitor cache...

// Please install OpenAI SDK first: `npm install openai` import OpenAI from...

* The deepseek-chat and deepseek-reasoner correspond to the model version...

Once you have obtained an API key, you can access the DeepSeek API using the...

openai/whisper-large-v3

Now : Field is deprecated; setting it has no effect (prompt caching is always on).

Use /audio/transcriptions and /audio/translations endpoint

Lower Cost: For most serverless models, the Batch Inference API runs at 50%...

Massive Scale Jump: Rate limits are up from 10M to 30B enqueued tokens per...

openai/gpt-oss-20b

DeepSeek-V3.1

Automatic filtering of sequences with no trainable tokens (e.g., if a...

Serverless Models

"safety_model": "MODEL_API_STRING"

openai/gpt-oss-20b

deepseek-ai/DeepSeek-V3.1

Pricing (per 1M tokens)

curl https://api.fireworks.ai/inference/v1/chat/completions \ -H...

Serverless model lifecycle

Rate Limits & Quotas

import Anthropic from "@anthropic-ai/sdk" ; const client = new Anthropic ({...

import os from openai import OpenAI client = OpenAI( api_key =...

Mistral Small 3 Now Available on Fireworks: Faster, Lighter, and More Efficient

Excited to launch a multi-year partnership bringing Fireworks to Microsoft...

Kimi K2.5 is Live on Fireworks: Vibe Coding, Agents, and Full-Parameter RFT

50 Trillion Tokens Per Day: The State of Agent Environments

Building a RAG with Astro, FastAPI, SurrealDB and Llama 3.1

Introducing Vision-Language Model Fine-tuning: Tailor VLMs to Your Domain

Fireworks launches fine-tuning service - Rapidly iterate on quality and...

GPUs on-demand: Not serverless, not reserved, but some third thing

$ 0.007800 /sec $ 28.08 /hr

Deploy a custom model

Explore the differences between Flux Schnell and Flux Dev image generation...

Learn how to receive webhooks from Replicate's API when running predictions...

Hello, world! We're a small team of engineers and machine learning...

Run FLUX with an API FLUX.1 is a new text-to-image model from Black Forest...

Wan2.1 parameter sweep We've been playing with Alibaba's WAN2.1...

Introducing LoRA: A faster way to fine-tune Stable Diffusion It's like...

Train and deploy a DreamBooth model on Replicate With just a handful of...

Explore the Sonar range and compare models

Models & Configuration

Rate Limits & Usage Tiers