Latest AI Model Releases

Black Forest Labs · Open-weight · released Oct 15, 2026 · Generally available

5.6B parameter open-weights image model

textimageopen-weight

Meta · Open-weight · released Oct 15, 2026 · Generally available

816B total / 42B active MoE

textimagevideoaudio1,000,000 contextopen-weight

OpenAI · Multimodal · released Oct 14, 2026 · Deprecated

Listed snapshots shut down 2027-01-14; recommended replacement gpt-6-luna

textimage

API id gpt-4o

Google · Flagship · released Oct 13, 2026 · Generally available

Google calls it a flagship reasoning model

textimagevideoaudiopdf1,048,576 context

API id gemini-4-argon · $2 in / $10 out per 1M tokens

Google · Fast / lightweight · released Oct 8, 2026 · Deprecated

Gemini 3.7 Flash replaced by 3.8 Flash; Gemini 3.5 Flash routed to 3.6 Flash

text

API id gemini-3.7-flash

Anthropic · Fast / lightweight · released Oct 7, 2026 · Generally available

1M token context window (Haiku 4.5 had less)

textimage1,000,000 context

API id claude-haiku-5-5 · $0.10 in / $0.50 out per 1M tokens

Mistral AI · Flagship · released Oct 6, 2026 · Preview

Natively multimodal MoE with about 1T total / 52B active parameters

textimage1,000,000 contextopen-weight

API id mistral-large-4 · $0.68 in / $2.09 out per 1M tokens

Google · Image · released Oct 6, 2026 · Generally available

Image generation and editing; gemini-3.1-flash-image deprecated the same day

textimagevideo

API id gemini-nano-banana-2.1 · $1.50 in / $30 out per 1M tokens

OpenAI · Speech / TTS · released Oct 1, 2026 · Deprecated

Shutdown 2027-01-06; recommended replacement gpt-realtime-2.1-mini

textaudio

API id tts-1

Anthropic · Fast / lightweight · released Sep 30, 2026 · Deprecated

Retirement on the Claude API 2026-11-30; replacement claude-sonnet-5-5

text

API id claude-sonnet-4-5-20250929

OpenAI · Reasoning · released Sep 29, 2026 · Generally available

Positioned as near-Astra performance at lower cost

textimage1,050,000 context

API id gpt-6.1-sol · $2 in / $10 out per 1M tokens

Anthropic · Fast / lightweight · released Sep 28, 2026 · Generally available

Anthropic says it is about 30% faster than Sonnet 5 and up to 30% cheaper for most work

textimage1,000,000 context

API id claude-sonnet-5-5 · $2 in / $10 out per 1M tokens

ElevenLabs · Speech / TTS · released Sep 28, 2026 · Generally available

ElevenLabs calls it its most emotive model

textaudio

API id eleven_v4

Anthropic · Flagship · released Sep 22, 2026 · Generally available

Anthropic says it performs at the level of Fable 5.1 on most work

textimage1,000,000 context

API id claude-opus-5-5 · $4 in / $20 out per 1M tokens

Google · Speech / TTS · released Sep 22, 2026 · Generally available

GA together with 3.8 Flash-Lite TTS and a Voices endpoint

textaudio

API id gemini-3.8-flash-tts

OpenAI · Fast / lightweight · released Sep 22, 2026 · Generally available

OpenAI calls it its most efficient model for focused, high-volume tasks

textimage1,050,000 context

API id gpt-6-luna · $0.10 in / $0.50 out per 1M tokens

OpenAI · Reasoning · released Sep 22, 2026 · Generally available

OpenAI says GPT-6 Sol and Luna cost half as much as the GPT-5.6 Sol and Luna models

textimage1,050,000 context

API id gpt-6-sol · $2 in / $10 out per 1M tokens

xAI · Flagship · released Sep 21, 2026 · Generally available

xAI docs call it the most capable model it has built, for coding, agentic and knowledge work

textimage500,000 context

API id grok-4.7 · $2 in / $6 out per 1M tokens

DeepSeek · Fast / lightweight · released Sep 10, 2026 · Generally available

Native multimodal support

textimage1,000,000 context

API id deepseek-flash · $0.30 in / $1.20 out per 1M tokens

OpenAI · Speech / TTS · released Sep 10, 2026 · Generally available

Expressive voice conversations with interruption handling

audiotext

API id gpt-live-1

Suno · Music · released Sep 9, 2026 · Generally available

Three variants: v6 and v6-wild (Pro/Premier) and v6-mini (free plan)

textaudio

OpenAI · Image · released Sep 8, 2026 · Generally available

Faster everyday image generation tier of GPT Image 2.5

textimage

API id gpt-image-2.5-flare

OpenAI · Image · released Sep 8, 2026 · Generally available

OpenAI calls it its most capable image generation and editing model

textimage

API id gpt-image-2.5-sunburst · $5 in / $30 out per 1M tokens

OpenAI · Flagship · released Sep 3, 2026 · Generally available

OpenAI describes it as its most capable model for demanding work

textimage1,050,000 context

API id gpt-6-astra · $10 in / $50 out per 1M tokens

Google · Fast / lightweight · released Sep 2, 2026 · Generally available

Google calls it its most intelligent Flash model

textimagevideoaudiopdf1,048,576 context

API id gemini-3.8-flash · $0.75 in / $3.75 out per 1M tokens

Meta · Flagship · released Sep 2, 2026 · Generally available

Trained for agentic workflows and competitive coding

textimagevideopdf1,000,000 context

API id muse-spark-1.3 · $1.25 in / $4.25 out per 1M tokens

Anthropic · Flagship · released Sep 1, 2026 · Generally available

Anthropic positions it for demanding reasoning and long-horizon agentic work

textimage1,000,000 context

API id claude-fable-5-1 · $10 in / $50 out per 1M tokens

Anthropic · Flagship · released Sep 1, 2026 · Generally available

Limited-availability counterpart of Fable 5.1 for selected organisations (per press, cybersecurity and life sciences)

textimage

API id claude-mythos-5-1 · $10 in / $50 out per 1M tokens

Google · Video · released Aug 27, 2026 · Generally available

Video generation and editing with native audio

textimagevideoaudio

API id gemini-omni-1.1-flash · $1.50 in / $9 out per 1M tokens

Alibaba (Qwen) · Fast / lightweight · released Aug 26, 2026 · Generally available

1M token context

text1,000,000 context

API id qwen3.8-flash

OpenAI · Transcription · released Aug 26, 2026 · Deprecated

whisper-1, gpt-4o-transcribe, gpt-4o-mini-transcribe, gpt-4o-transcribe-diarize and /v1/audio/translations shut down 2027-02-26

audiotext

API id whisper-1

DeepSeek · Flagship · released Aug 13, 2026 · Generally available

V4 first available 2026-04-24; GA as deepseek-v4-pro on 2026-08-13 (snapshot 0813)

text1,000,000 context

API id deepseek-v4-pro · $1.32 in / $3.96 out per 1M tokens

Alibaba (Qwen) · Open-weight · released Aug 13, 2026 · Generally available

Open-weight version of the Qwen3.8 flagship: 2.4T total / ~95B active

textimage1,000,000 contextopen-weight

Alibaba (Qwen) · Flagship · released Aug 3, 2026 · Generally available

2.4T-parameter MoE

textimage1,000,000 context

API id qwen3.8-max · $2 in / $6 out per 1M tokens

OpenAI · Transcription · released Jul 28, 2026 · Generally available

Supports context, keyword hints and multiple language hints

audiotext

API id gpt-transcribe

Midjourney · Image · released Jul 24, 2026 · Generally available

Became the default model on 2026-07-24

textimage

Black Forest Labs · Video · released Jul 23, 2026 · Waitlist only

Multimodal family: FLUX 3 Video (up to 20s with audio), FLUX 3 Image, FLUX 3 Audio and FLUX 3 Action (open weights)

textimagevideoaudio

Google · Fast / lightweight · released Jul 21, 2026 · Generally available

Fastest model of the 3.5 generation

textimagevideoaudio

API id gemini-3.5-flash-lite · $0.30 in / $2.50 out per 1M tokens

OpenAI · Flagship · released Jul 9, 2026 · Generally available

Launched alongside GPT-5.6 Terra and GPT-5.6 Luna as a three-tier family

textimage1,050,000 context

API id gpt-5.6-sol · $4 in / $20 out per 1M tokens

Google · Image · released Jun 15, 2026 · Discontinued

Shutdown announced 2026-06-15 for 2026-08-17; models page lists imagen-4.0-generate as shut down

textimage

API id imagen-4.0-generate

OpenAI · Reasoning · released Jun 11, 2026 · Deprecated

gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, o3 and o3-pro snapshots shut down 2026-12-11

text

API id o3

Cohere · Flagship · released May 20, 2026 · Generally available

Cohere's first MoE model (218B total)

textimage128,000 contextopen-weight

API id command-a-plus-05-2026

Mistral AI · Multimodal · released Apr 28, 2026 · Generally available

Multimodal with adjustable reasoning via reasoning_effort

textimageopen-weight

API id mistral-medium-3-5 · $1.50 in / $7.50 out per 1M tokens

DeepSeek · Reasoning · released Apr 24, 2026 · Deprecated

Scheduled for discontinuation 2026-07-24; meanwhile mapped to V4-Flash non-thinking and thinking modes

text

API id deepseek-reasoner

Google · Embedding · released Apr 22, 2026 · Generally available

Multimodal embeddings across text, images, video, audio and PDFs (preview released 2026-03-10)

textimagevideoaudiopdf

API id gemini-embedding-2 · $0.20 in / out per 1M tokens

Google DeepMind · Open-weight · released Mar 31, 2026 · Generally available

Sizes E2B, E4B, 26B A4B (MoE) and 31B dense

textimageopen-weight

OpenAI · Video · released Mar 24, 2026 · Deprecated

Videos API and sora-2 / sora-2-pro snapshots scheduled to shut down 2026-09-24; no replacement listed

textimagevideoaudio

API id sora-2

Mistral AI · Fast / lightweight · released Mar 16, 2026 · Generally available

Hybrid instruct, reasoning and coding model

text256,000 contextopen-weight

API id mistral-small-2603 · $0.15 in / $0.60 out per 1M tokens

Google · Video · released Oct 15, 2025 · Preview

Video generation with synchronized audio

textimagevideoaudio

API id veo-3.1-generate-preview

Mistral AI · Coding · release date not verified · Generally available

Current Codestral version listed in the models overview; no 2026 Codestral/Devstral release

text

$0.30 in / $0.90 out per 1M tokens

Cohere · Embedding · release date not verified · Generally available

128K context

textimage128,000 context

API id embed-v5.0-pro

xAI · Speech / TTS · release date not verified · Generally available

Adds speech-to-speech at the same price as Voice 2.0; grok-voice-latest now points to it; Voice 2.0 deprecated

audiotext

API id grok-voice-3.0

Meta · Open-weight · release date not verified · Generally available

30B model distilled from Muse Spark, built for local agents

textimageopen-weight