o3
o3Deep-thinking reasoning model, extended chain-of-thought for hard problems.
FrontierReasoningCodingLong context
- Context
- 200K tokens
- Released
- 2025-04
- Input
- $10 / 1M tokens
- Output
- $40 / 1M tokens
- Cached
- $2.5 / 1M tokens
The index
The aiDex.
Every mind, with their strengths. Update yourself before you choose your team.
Output
Capability
Provider
57 models
OpenAI
The original. Strong all-rounders plus deep-thinking reasoning models.
GPT-5.4
gpt-5.4OpenAI's flagship, best for complex reasoning and code.
FrontierReasoningCodingVisionLong context
- Context
- 400K tokens
- Released
- 2025-12
- Input
- $2.5 / 1M tokens
- Output
- $15 / 1M tokens
- Cached
- $1.25 / 1M tokens
GPT-4o
gpt-4oMultimodal generalist, strong all-rounder.
ReasoningCodingVisionMultilingual
- Context
- 128K tokens
- Released
- 2024-05
- Input
- $2.5 / 1M tokens
- Output
- $10 / 1M tokens
- Cached
- $1.25 / 1M tokens
GPT-5.4 mini
gpt-5.4-miniCheaper, faster GPT-5.4 for everyday work.
FastCodingWriting
- Context
- 400K tokens
- Released
- 2025-12
- Input
- $0.75 / 1M tokens
- Output
- $4.5 / 1M tokens
- Cached
- $0.375 / 1M tokens
o3-mini
o3-miniFast reasoning, great for math, code, and logic.
ReasoningCodingFast
- Context
- 200K tokens
- Released
- 2025-01
- Input
- $1.1 / 1M tokens
- Output
- $4.4 / 1M tokens
- Cached
- $0.55 / 1M tokens
GPT-4o mini
gpt-4o-miniCheapest OpenAI option, great for high volume.
FastWriting
- Context
- 128K tokens
- Released
- 2024-07
- Input
- $0.15 / 1M tokens
- Output
- $0.6 / 1M tokens
- Cached
- $0.075 / 1M tokens
GPT Image 1
gpt-image-1OpenAI's flagship image model, strong prompt adherence, three quality tiers and aspect ratios.
Vision
- Released
- 2025-04
- Price
- $0.042 / image
DALLĀ·E 3
dall-e-3Legacy OpenAI image model, kept for orgs without GPT Image 1 access yet.
Vision
- Released
- 2023-10
- Price
- $0.04 / image
Sora 2
sora-2OpenAI's text-to-video, short cinematic clips up to ~20s with synced audio.
FrontierVision
- Released
- 2025-09
Price varies by host
o1
o1Previous-generation reasoning flagship, still capable, broadly available.
ReasoningCodingLong context
- Context
- 200K tokens
- Released
- 2024-12
- Input
- $15 / 1M tokens
- Output
- $60 / 1M tokens
- Cached
- $7.5 / 1M tokens
GPT-4 Turbo
gpt-4-turboLegacy GPT-4 with 128K context, kept for compatibility.
CodingWritingLong context
- Context
- 128K tokens
- Released
- 2024-04
- Input
- $10 / 1M tokens
- Output
- $30 / 1M tokens
- Cached
- $5 / 1M tokens
Anthropic
Best-in-class at code, careful writing, and nuanced analysis.
Claude Opus 4.8
claude-opus-4.8Anthropic's most capable model, complex reasoning and agentic coding.
FrontierReasoningCodingWritingLong context
- Context
- 1M tokens
- Released
- 2026-05
- Input
- $5 / 1M tokens
- Output
- $25 / 1M tokens
- Cached
- $0.5 / 1M tokens
Claude Opus 4.7
claude-opus-4.7Anthropic's deepest model, nuanced writing and analysis.
FrontierReasoningWritingLong context
- Context
- 1M tokens
- Released
- 2026-01
- Input
- $5 / 1M tokens
- Output
- $25 / 1M tokens
- Cached
- $0.5 / 1M tokens
Claude Sonnet 4.6
claude-sonnet-4.6Balanced flagship, especially good at code.
FrontierCodingWritingReasoningLong context
- Context
- 200K tokens
- Released
- 2025-11
- Input
- $3 / 1M tokens
- Output
- $15 / 1M tokens
- Cached
- $0.3 / 1M tokens
Claude Haiku 3.5
claude-haiku-3.5Previous-generation Haiku, still capable, lower cost.
FastWritingCoding
- Context
- 200K tokens
- Released
- 2024-11
- Input
- $0.8 / 1M tokens
- Output
- $4 / 1M tokens
- Cached
- $0.08 / 1M tokens
Claude Haiku 4.5
claude-haiku-4.5Cheapest and fastest Claude, solid default.
FastWriting
- Context
- 200K tokens
- Released
- 2025-10
- Input
- $0.25 / 1M tokens
- Output
- $1.25 / 1M tokens
- Cached
- $0.025 / 1M tokens
Claude 3.5 Sonnet
claude-3-5-sonnetLegacy Sonnet, strong general-purpose model.
CodingWritingReasoning
- Context
- 200K tokens
- Released
- 2024-10
- Input
- $3 / 1M tokens
- Output
- $15 / 1M tokens
- Cached
- $0.3 / 1M tokens
Huge context windows, native vision, and strong multilingual support.
Gemini 3.1 Pro
gemini-3.1-proGoogle's flagship, huge context window and strong vision.
FrontierReasoningVisionLong contextMultilingual
- Context
- 2M tokens
- Released
- 2025-12
- Input
- $2 / 1M tokens
- Output
- $12 / 1M tokens
- Cached
- $1 / 1M tokens
Gemini 2.5 Pro
gemini-2.5-proPrevious-gen Gemini Pro, still very capable.
ReasoningVisionLong contextMultilingual
- Context
- 2M tokens
- Released
- 2025-03
- Input
- $1.25 / 1M tokens
- Output
- $7 / 1M tokens
- Cached
- $0.31 / 1M tokens
Gemini 3 Flash
gemini-3-flashFast and cheap Gemini for high-volume tasks.
FastVisionMultilingual
- Context
- 1M tokens
- Released
- 2025-11
- Input
- $0.5 / 1M tokens
- Output
- $3 / 1M tokens
- Cached
- $0.25 / 1M tokens
Gemini 2.5 Flash
gemini-2.5-flashPrevious-gen Flash, cheap and dependable.
FastMultilingual
- Context
- 1M tokens
- Released
- 2025-06
- Input
- $0.3 / 1M tokens
- Output
- $1.2 / 1M tokens
- Cached
- $0.075 / 1M tokens
Nano Banana Pro
gemini-3-pro-imageGemini 3 Pro Image, top-tier fidelity, photoreal lighting, slower and pricier than Flash.
FrontierVision
- Released
- 2025-12
- Price
- $0.12 / image
Nano Banana 2
gemini-3.1-flash-imageGemini 3.1 Flash Image, next-gen Flash, sharper than 2.5, cheaper than Pro.
VisionFast
- Released
- 2026-03
- Price
- $0.06 / image
Nano Banana
gemini-2.5-flash-imageGemini 2.5 Flash Image, fast, cheap and surprisingly sharp. Flat per-image price.
VisionFast
- Released
- 2025-08
- Price
- $0.039 / image
Gemini 1.5 Pro
gemini-1.5-proOlder Pro with 2M context, kept for awareness.
ReasoningVisionLong context
- Context
- 2M tokens
- Released
- 2024-05
- Input
- $1.25 / 1M tokens
- Output
- $5 / 1M tokens
- Cached
- $0.313 / 1M tokens
Veo 3
veo-3Google's video model, 4K output, synchronised audio, strong physics.
FrontierVision
- Released
- 2025-05
- Price
- $0.5 / second
Veo 2
veo-2Previous-gen Veo, still capable for short B-roll generation.
Vision
- Released
- 2024-12
- Price
- $0.35 / second
Gemini 1.5 Flash
gemini-1.5-flashOlder Flash, cheaper but no longer recommended.
Fast
- Context
- 1M tokens
- Released
- 2024-05
- Input
- $0.075 / 1M tokens
- Output
- $0.3 / 1M tokens
- Cached
- $0.019 / 1M tokens
DeepSeek
Open-weight reasoning at a fraction of frontier-lab prices.
DeepSeek R1
deepseek-r1Open-weight reasoning model, competitive with o-series at a much lower price.
ReasoningCodingLong context
- Context
- 64K tokens
- Released
- 2025-01
- Input
- $0.55 / 1M tokens
- Output
- $2.19 / 1M tokens
- Cached
- $0.14 / 1M tokens
DeepSeek V3.2
deepseek-v3.2Strong at code and reasoning at a fraction of the cost.
CodingReasoningFast
- Context
- 64K tokens
- Released
- 2025-09
- Input
- $0.14 / 1M tokens
- Output
- $0.28 / 1M tokens
- Cached
- $0.014 / 1M tokens
Ollama
Open-weight models running on your own hardware. Zero cost per call.
Llama 3.1 8B
llama3.1:8bRuns locally via Ollama, free, private, offline-capable.
Open-weightFast
- Context
- 128K tokens
- Released
- 2024-07
FreeRuns on your own machine via Ollama. No API bill, never leaves your computer.
Qwen 2.5 7B
qwen2.5:7bOpen-weight Qwen, strong at code and non-English.
Open-weightCodingMultilingual
- Context
- 128K tokens
- Released
- 2024-09
FreeRuns on your own machine via Ollama. No API bill, never leaves your computer.
Mistral 7B
mistral:7bOpen-weight Mistral, efficient generalist.
Open-weightFastMultilingual
- Context
- 32K tokens
- Released
- 2023-09
FreeRuns on your own machine via Ollama. No API bill, never leaves your computer.
Phi-3 14B
phi3:14bMicrosoft's open-weight Phi-3, strong at reasoning despite small size.
Open-weightReasoningFast
- Context
- 128K tokens
- Released
- 2024-04
FreeRuns on your own machine via Ollama. No API bill, never leaves your computer.
Mixtral 8x7B
mixtral:8x7bMoE model from Mistral, fast inference with broad coverage.
Open-weightCodingFastMultilingual
- Context
- 32K tokens
- Released
- 2023-12
FreeRuns on your own machine via Ollama. No API bill, never leaves your computer.
DeepSeek Coder 6.7B
deepseek-coder:6.7bCode-specialised open-weight model, runs on modest hardware.
Open-weightCodingFast
- Context
- 16K tokens
- Released
- 2024-01
FreeRuns on your own machine via Ollama. No API bill, never leaves your computer.
Gemma 2 9B
gemma2:9bGoogle's open-weight family, solid writing and reasoning.
Open-weightWritingFast
- Context
- 8K tokens
- Released
- 2024-06
FreeRuns on your own machine via Ollama. No API bill, never leaves your computer.
xAI
xAI's lab, real-time web access and a candid tone.
Grok 4
Coming soongrok-4xAI's flagship reasoning model, real-time web access and strong on long-context tasks.
FrontierReasoningLong context
- Context
- 256K tokens
- Released
- 2025-07
- Input
- $3 / 1M tokens
- Output
- $15 / 1M tokens
- Cached
- $0.75 / 1M tokens
Grok 2
Coming soongrok-2Previous-gen Grok, capable generalist with a wry tone.
ReasoningWriting
- Context
- 131K tokens
- Released
- 2024-08
- Input
- $2 / 1M tokens
- Output
- $10 / 1M tokens
- Cached
- $0.5 / 1M tokens
Mistral AI
European frontier lab. Strong code models and open weights.
Mistral Large 2
Coming soonmistral-large-2Mistral's hosted flagship, strong code and multilingual.
FrontierCodingReasoningMultilingual
- Context
- 128K tokens
- Released
- 2024-07
- Input
- $2 / 1M tokens
- Output
- $6 / 1M tokens
- Cached
- $1 / 1M tokens
Mistral Medium 3
Coming soonmistral-medium-3Mid-tier hosted Mistral, good cost-to-quality ratio.
CodingWritingFast
- Context
- 32K tokens
- Released
- 2025-05
- Input
- $0.4 / 1M tokens
- Output
- $2 / 1M tokens
- Cached
- $0.2 / 1M tokens
Codestral
Coming sooncodestralCode-specialised Mistral, autocomplete + chat for 80+ languages.
CodingFast
- Context
- 32K tokens
- Released
- 2024-05
- Input
- $0.3 / 1M tokens
- Output
- $0.9 / 1M tokens
- Cached
- $0.15 / 1M tokens
Cohere
Enterprise-focused, RAG-tuned models with strong citations and tool use.
Command R+
Coming sooncommand-r-plusCohere's RAG-tuned flagship, strong on retrieval, citations, and tool use.
ReasoningWritingLong contextMultilingual
- Context
- 128K tokens
- Released
- 2024-04
- Input
- $2.5 / 1M tokens
- Output
- $10 / 1M tokens
- Cached
- $1.25 / 1M tokens
Command R
Coming sooncommand-rCheaper Command, same RAG strengths, smaller envelope.
WritingFastMultilingual
- Context
- 128K tokens
- Released
- 2024-03
- Input
- $0.15 / 1M tokens
- Output
- $0.6 / 1M tokens
- Cached
- $0.075 / 1M tokens
Meta
The Llama family. Open weights, frontier-tier, run anywhere.
Llama 3.3 70B
llama-3.3-70bMeta's open-weight flagship, runs locally on beefy hardware via Ollama, or cloud via Together / Replicate.
Open-weightFrontierReasoningCodingLong context
- Context
- 128K tokens
- Released
- 2024-12
Price varies by host
Llama 4 Maverick
llama-4-maverickNext-gen Llama from Meta, flagship reasoning, multimodal, open-weight.
Open-weightFrontierReasoningVision
- Context
- 1M tokens
- Released
- 2025-04
Price varies by host
Perplexity
Web-grounded models, answers with live citations.
Sonar Pro
sonar-proPerplexity's web-grounded model, answers backed by live search. Different shape than a plain chat completion.
ReasoningWriting
- Context
- 200K tokens
- Released
- 2025-01
- Input
- $3 / 1M tokens
- Output
- $15 / 1M tokens
- Cached
- $0.75 / 1M tokens
MidJourney
Distinctive aesthetic image model, favourite among designers and art directors.
MidJourney v7
midjourney-v7MidJourney's flagship, distinctive aesthetic, no public API yet (Discord / web-only).
FrontierVision
- Released
- 2025-04
Price varies by host
MidJourney v6.1
midjourney-v6.1Previous-gen MidJourney, still the photographer's favourite.
Vision
- Released
- 2024-07
Price varies by host
Black Forest Labs
FLUX models from the team behind Stable Diffusion, sharp prompt following, fast.
FLUX 1.1 Pro
Coming soonflux-1.1-proBlack Forest Labs' flagship, sharp prompt following, fast generation, public API.
FrontierVision
- Released
- 2024-10
- Price
- $0.04 / image
FLUX 1 Schnell
Coming soonflux-1-schnellOpen-weight FLUX, runs locally on consumer GPUs, free to use.
Open-weightVisionFast
- Released
- 2024-08
FreeRuns on your own machine via Ollama. No API bill, never leaves your computer.
Stability AI
Stable Diffusion family, open weights, vast community tooling.
Stable Diffusion 3.5
stable-diffusion-3.5Open-weight SD3.5, runs locally, hugely customisable via LoRAs and ControlNet.
Open-weightVision
- Released
- 2024-10
Price varies by host
SDXL
sdxlClassic Stable Diffusion XL, broad community + tooling support.
Open-weightVision
- Released
- 2023-07
Price varies by host
Runway
Hosted video generation, character consistency and scene control for filmmakers.
Runway Gen-4
runway-gen-4Runway's next-gen video, character consistency, scene control, image-to-video.
FrontierVision
- Released
- 2025-03
Price varies by host
Runway Gen-3 Alpha
runway-gen-3-alphaRunway's previous flagship, fast, hosted, popular with filmmakers.
Vision
- Released
- 2024-07
Price varies by host
Pika
Friendly, fast video generation, image-to-video and scene-extend.
Pika 2.0
pika-2.0Pika's 2.0, image-to-video and scene-extend, friendly UX, fast turnaround.
VisionFast
- Released
- 2024-12
Price varies by host
Kling
Kuaishou's video lab, long-form clips and strong human motion.
Kling 2.0
kling-2.0Kuaishou's Kling, long-form video (up to 2 min), strong human motion and Asian-language prompts.
FrontierVisionMultilingual
- Released
- 2025-04
Price varies by host
Prices are USD, sourced directly from each provider's rate card, per million tokens for chat models, per image for image models, per second for video models. On BYOK you pay each provider directly with no markup from us; on Pro and Power we cover the provider cost from your Monthly AI Credits. "Coming soon" entries are on our roadmap; "Listed" entries are known to exist but not currently planned for direct integration.