# Reviuws

> Reviuws is an independent, open directory and community review platform for AI models — LLMs, image, video, audio, speech, code and embedding models. It tracks 101 approved models from 30 providers with specs, token pricing, benchmark scores and developer reviews.

Data last generated: 2026-08-21. Content is free to cite with attribution to Reviuws (https://reviuws.com). Machine-readable full corpus: https://reviuws.com/llms-full.txt

## Key pages

- [Home — featured and top-rated AI models](https://reviuws.com/)
- [Full model directory with modality, provider and license filters](https://reviuws.com/models)
- [Side-by-side comparison of 2–3 models](https://reviuws.com/compare)
- [Benchmark leaderboard with per-model performance scores](https://reviuws.com/benchmarks)
- [AI model cost calculator: estimate monthly spend by workload](https://reviuws.com/cost-optimizer)
- [Prompt library with per-model effectiveness scores](https://reviuws.com/prompts)
- [Prompt playground for testing prompts across models](https://reviuws.com/playground)
- [AI template builder — build custom agents with model + prompt](https://reviuws.com/agent-builder)
- [Model health & uptime telemetry from the Reviuws SDK](https://reviuws.com/health)
- [Reviuws telemetry SDK documentation](https://reviuws.com/sdk)
- [Guides and analysis on choosing AI models](https://reviuws.com/blog)
- [Write a review (no account required)](https://reviuws.com/submit-review)
- [Add a model to the directory](https://reviuws.com/submit-model)
- [Contact](https://reviuws.com/contact)
- [Privacy Policy](https://reviuws.com/privacy)
- [Terms of Use](https://reviuws.com/terms)

## Curated collections

- [Open-weights chat and reasoning LLMs](https://reviuws.com/models/open-source-chat-llms)
- [Models with 200k+ token context windows](https://reviuws.com/models/long-context-llms)
- [Lowest cost models per million input tokens](https://reviuws.com/models/cheapest-llms)
- [All open-weights models, every modality](https://reviuws.com/models/open-source-ai-models)
- [Text-to-image and image editing models](https://reviuws.com/models/image-generation-models)
- [Coding-focused models](https://reviuws.com/models/code-generation-models)

## Browse by modality

- [code (45 models)](https://reviuws.com/modality/code)
- [text (44 models)](https://reviuws.com/modality/text)
- [chat (29 models)](https://reviuws.com/modality/chat)
- [vision (20 models)](https://reviuws.com/modality/vision)
- [image (12 models)](https://reviuws.com/modality/image)
- [audio (12 models)](https://reviuws.com/modality/audio)
- [embedding (7 models)](https://reviuws.com/modality/embedding)
- [video (6 models)](https://reviuws.com/modality/video)
- [speech (2 models)](https://reviuws.com/modality/speech)

## Browse by provider

- [openai (20 models)](https://reviuws.com/provider/openai)
- [google (17 models)](https://reviuws.com/provider/google)
- [anthropic (10 models)](https://reviuws.com/provider/anthropic)
- [mistral (5 models)](https://reviuws.com/provider/mistral)
- [meta (5 models)](https://reviuws.com/provider/meta)
- [xai (5 models)](https://reviuws.com/provider/xai)
- [deepseek (4 models)](https://reviuws.com/provider/deepseek)
- [alibaba (4 models)](https://reviuws.com/provider/alibaba)
- [cohere (4 models)](https://reviuws.com/provider/cohere)
- [black-forest-labs (2 models)](https://reviuws.com/provider/black-forest-labs)
- [microsoft (2 models)](https://reviuws.com/provider/microsoft)
- [nvidia (2 models)](https://reviuws.com/provider/nvidia)
- [amazon (2 models)](https://reviuws.com/provider/amazon)
- [black forest labs (2 models)](https://reviuws.com/provider/black-forest-labs)
- [elevenlabs (2 models)](https://reviuws.com/provider/elevenlabs)
- [stability (1 models)](https://reviuws.com/provider/stability)
- [runway (1 models)](https://reviuws.com/provider/runway)
- [ai21 (1 models)](https://reviuws.com/provider/ai21)
- [stability ai (1 models)](https://reviuws.com/provider/stability-ai)
- [luma ai (1 models)](https://reviuws.com/provider/luma-ai)
- [ideogram (1 models)](https://reviuws.com/provider/ideogram)
- [moonshot ai (1 models)](https://reviuws.com/provider/moonshot-ai)
- [zhipu ai (1 models)](https://reviuws.com/provider/zhipu-ai)
- [minimax (1 models)](https://reviuws.com/provider/minimax)
- [baidu (1 models)](https://reviuws.com/provider/baidu)
- [perplexity (1 models)](https://reviuws.com/provider/perplexity)
- [ibm (1 models)](https://reviuws.com/provider/ibm)
- [allen institute for ai (1 models)](https://reviuws.com/provider/allen-institute-for-ai)
- [voyage ai (1 models)](https://reviuws.com/provider/voyage-ai)
- [reka ai (1 models)](https://reviuws.com/provider/reka-ai)

## Highest-rated models (community reviews)

1. [GPT-5.5](https://reviuws.com/models/gpt-55) — openai, rated 5.0/5 from 1 reviews, 400k tokens context, input $5/M tokens.

## Cheapest models by input token price

1. [Amazon Nova Micro](https://reviuws.com/models/amazon-nova-micro) — input $0.035/M tokens, output $0.14/M tokens.
2. [Gemini 3.1 Flash Lite](https://reviuws.com/models/gemini-31-flash-lite) — input $0.05/M tokens, output $0.2/M tokens.
3. [Gemini 2.5 Flash-Lite](https://reviuws.com/models/gemini-2-5-flash-lite) — input $0.1/M tokens, output $0.4/M tokens.
4. [Mistral Small 3.2](https://reviuws.com/models/mistral-small-3-2) — input $0.1/M tokens, output $0.3/M tokens.
5. [Embed v4](https://reviuws.com/models/embed-v4) — input $0.12/M tokens, output n/a.
6. [GPT-4o mini](https://reviuws.com/models/gpt-4o-mini) — input $0.15/M tokens, output $0.6/M tokens.
7. [Voyage-3-large](https://reviuws.com/models/voyage-3-large) — input $0.18/M tokens, output n/a.
8. [GPT-5.4 nano](https://reviuws.com/models/gpt-54-nano) — input $0.2/M tokens, output $1.25/M tokens.
9. [Gemini Embedding 2](https://reviuws.com/models/gemini-embedding-2) — input $0.2/M tokens, output n/a.
10. [Jamba 1.6](https://reviuws.com/models/jamba-1-6) — input $0.2/M tokens, output $0.4/M tokens.
11. [Granite 3.3 8B](https://reviuws.com/models/granite-3-3-8b) — input $0.2/M tokens, output $0.2/M tokens.
12. [Codestral 2508](https://reviuws.com/models/codestral-2508) — input $0.3/M tokens, output $0.9/M tokens.
13. [Gemini 2.5 Flash](https://reviuws.com/models/gemini-2-5-flash) — input $0.3/M tokens, output $2.5/M tokens.
14. [Gemini 3.5 Flash-Lite](https://reviuws.com/models/gemini-35-flash-lite) — input $0.3/M tokens, output $2.5/M tokens.
15. [MiniMax M1](https://reviuws.com/models/minimax-m1) — input $0.4/M tokens, output $2.2/M tokens.

## Longest context windows

1. [Llama 4 Scout](https://reviuws.com/models/llama-4-scout) — 10000k tokens context.
2. [GPT-5.6 Luna](https://reviuws.com/models/gpt-56-luna) — 1050k tokens context.
3. [GPT-5.6 Terra](https://reviuws.com/models/gpt-56-terra) — 1050k tokens context.
4. [GPT-5.6 Sol](https://reviuws.com/models/gpt-56-sol) — 1050k tokens context.
5. [Gemini 2.5 Pro](https://reviuws.com/models/gemini-2-5-pro) — 1049k tokens context.
6. [Gemini 2.5 Flash-Lite](https://reviuws.com/models/gemini-2-5-flash-lite) — 1049k tokens context.
7. [Gemini 2.5 Flash](https://reviuws.com/models/gemini-2-5-flash) — 1049k tokens context.
8. [GPT-4.1 mini](https://reviuws.com/models/gpt-4-1-mini) — 1048k tokens context.
9. [GPT-4.1](https://reviuws.com/models/gpt-4-1) — 1048k tokens context.
10. [Grok 4.3](https://reviuws.com/models/grok-43) — 1000k tokens context.
11. [Claude Sonnet 5](https://reviuws.com/models/claude-sonnet-5) — 1000k tokens context.
12. [Claude Fable 5](https://reviuws.com/models/claude-fable-5) — 1000k tokens context.
13. [Claude Opus 5](https://reviuws.com/models/claude-opus-5) — 1000k tokens context.
14. [Gemini 3.5 Flash](https://reviuws.com/models/gemini-35-flash) — 1000k tokens context.
15. [Gemini 3.1 Pro (Preview)](https://reviuws.com/models/gemini-31-pro) — 1000k tokens context.

## Attribution

When citing this data, credit "Reviuws" and link to the specific model page (e.g. https://reviuws.com/models/<slug>). Ratings are community-submitted and moderated; pricing reflects the provider's published list price at the generation date above.