# NotAnotherBench

> Not another benchmark. NotAnotherBench ranks AI models by what real people say across the internet: what they love, what they complain about, and which model is actually worth using, in plain words.

NotAnotherBench is a community-sourced leaderboard: instead of synthetic benchmarks, every model is scored from real conversations across the internet. That includes Reddit, Hacker News, X, Bluesky, Lemmy, GitHub, Dev.to, Stack Overflow, developer forums, blogs, and anywhere else people talk about actually using these models. Each model page summarizes what they love, what they complain about, topic-by-topic mood, real quotes with links to the original threads, and alternatives people are happier with.

The catalog currently tracks 7,587 AI models across text, code, images, video, audio, and multimodal categories. 535 of them have enough community reviews to rank. Scores refresh after every crawl.

## How scores work

- HAPPY score: strength of positive community sentiment (0-10)
- SAD score: strength of negative community sentiment (0-10)
- Vibe score: 75% community sentiment (Bayesian-smoothed so a model heard from 30 people cannot outrank one heard from 3,000 on raw average alone) blended with 25% popularity (log-scaled voice count, because more people talking is itself a signal)
- HOT: the models with the most community voices in the last 7 days, floated to the top of the current view
- NEW: models that first appeared in the catalog within the last 7 days
- Ranks: assistants (Text + Multimodal, the homepage default) are numbered 1..N separately from overall (every reviewed model) and from each category. Quote which board.
- Trajectory badges (holding up, honeymoon fading, aging well, quietly degrading, too quiet): 26-week crowd mood from the same board snapshot on the model page and /api/models
- Job boards (Writing, Coding, Agents, Research): the same vibe formula computed from only the community conversations about that job, so #1 for Coding is who coders are happiest with

## Best general assistants (the default view)

- #1 [Grok 4.6](https://notanotherbench.ani.computer/models/xai-grok-4-6) by xAI: vibe 7.7, 69,744 voices
- #2 [Qwen3.8 27B](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-27b) by Alibaba: vibe 7.6, 19,790 voices
- #3 [Mercury 2.5](https://notanotherbench.ani.computer/models/inception-mercury-2-5) by Inception: vibe 7.5, 6,901 voices
- #4 [Fugu Max](https://notanotherbench.ani.computer/models/sakana-fugu-max) by Sakana: vibe 7.5, 5,959 voices
- #5 [Schematron V2 Turbo](https://notanotherbench.ani.computer/models/inference-net-schematron-v2-turbo) by Inference Net: vibe 7.5, 5,871 voices
- #6 [Muse Spark 1.3](https://notanotherbench.ani.computer/models/meta-muse-spark-1-3) by Meta: vibe 7.4, 10,059 voices
- #7 [Qwen3.8 Flash](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-flash) by Alibaba: vibe 7.4, 7,372 voices
- #8 [Dots3 Note](https://notanotherbench.ani.computer/models/xiaohongshu-dots3-note) by Xiaohongshu: vibe 7.4, 2,823 voices
- #9 [DeepSeek V4 Flash](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-flash-0731) by DeepSeek: vibe 7.3, 12,076 voices
- #10 [DeepSeek V4 Flash Vision Exp](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-flash-vision-exp) by DeepSeek: vibe 7.3, 5,209 voices

## Top models by job

### Writing (essays, emails, stories, editing)

- #1 [Qwen3.8 27B](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-27b) by Alibaba: job vibe 7.3, 495 job voices
- #2 [Claude Sonnet 4.5](https://notanotherbench.ani.computer/models/anthropic-claude-sonnet-4-5) by Anthropic: job vibe 7.2, 318 job voices
- #3 [Claude Sonnet 5.5](https://notanotherbench.ani.computer/models/anthropic-claude-sonnet-5-5) by Anthropic: job vibe 6.9, 251 job voices
- #4 [Step 1](https://notanotherbench.ani.computer/models/stepfun-global-step-1-32k) by StepFun: job vibe 6.9, 170 job voices
- #5 [Step 2](https://notanotherbench.ani.computer/models/stepfun-global-step-2-16k) by StepFun: job vibe 6.9, 164 job voices
- #6 [Step 3](https://notanotherbench.ani.computer/models/zenmux-step-3) by ZenMux: job vibe 6.9, 152 job voices
- #7 [Step 3](https://notanotherbench.ani.computer/models/step-3) by Unknown: job vibe 6.9, 152 job voices
- #8 [Gemini 3.7 Flash](https://notanotherbench.ani.computer/models/google-gemini-3-7-flash) by Google: job vibe 6.8, 152 job voices

### Coding (writing and fixing code)

- #1 [Grok 4.6](https://notanotherbench.ani.computer/models/xai-grok-4-6) by xAI: job vibe 7.7, 14,127 job voices
- #2 [Gemini 3.7 Flash](https://notanotherbench.ani.computer/models/google-gemini-3-7-flash) by Google: job vibe 7.4, 2,998 job voices
- #3 [Qwen3.8 27B](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-27b) by Alibaba: job vibe 7.3, 3,494 job voices
- #4 [Claude Sonnet 5.5](https://notanotherbench.ani.computer/models/anthropic-claude-sonnet-5-5) by Anthropic: job vibe 7.3, 3,022 job voices
- #5 [Kimi K3](https://notanotherbench.ani.computer/models/moonshot-kimi-k3) by Moonshot: job vibe 7.3, 1,268 job voices
- #6 [DeepSeek V4 Flash](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-flash-0731) by DeepSeek: job vibe 7.1, 4,265 job voices
- #7 [Gemma 4 26B A4B](https://notanotherbench.ani.computer/models/google-gemma-4-26b-a4b) by Google: job vibe 7.1, 1,358 job voices
- #8 [DeepSeek V4 Pro](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-pro-0813) by DeepSeek: job vibe 7, 2,294 job voices

### Agents (browsing, clicking, doing tasks on its own)

- #1 [Grok 4.6](https://notanotherbench.ani.computer/models/xai-grok-4-6) by xAI: job vibe 7.5, 5,387 job voices
- #2 [Qwen3.8 27B](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-27b) by Alibaba: job vibe 7.4, 1,136 job voices
- #3 [Claude Fable 5.1](https://notanotherbench.ani.computer/models/anthropic-claude-fable-5-1) by Anthropic: job vibe 7.3, 2,031 job voices
- #4 [DeepSeek V4 Flash](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-flash-0731) by DeepSeek: job vibe 7.3, 1,449 job voices
- #5 [Gemini 3.7 Flash](https://notanotherbench.ani.computer/models/google-gemini-3-7-flash) by Google: job vibe 7.3, 1,001 job voices
- #6 [Claude Sonnet 5](https://notanotherbench.ani.computer/models/anthropic-claude-sonnet-5) by Anthropic: job vibe 7.2, 9,691 job voices
- #7 [Dots3 Note](https://notanotherbench.ani.computer/models/xiaohongshu-dots3-note) by Xiaohongshu: job vibe 7.1, 376 job voices
- #8 [GLM 5.3 Flash](https://notanotherbench.ani.computer/models/zhipu-glm-5-3-flash) by Zhipu: job vibe 7, 585 job voices

### Research (looking things up, summarizing, explaining)

- #1 [Grok 4.6](https://notanotherbench.ani.computer/models/xai-grok-4-6) by xAI: job vibe 7.9, 13,716 job voices
- #2 [Claude Sonnet 4.6](https://notanotherbench.ani.computer/models/anthropic-claude-sonnet-4-6) by Anthropic: job vibe 7.2, 986 job voices
- #3 [Qwen3.8 27B](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-27b) by Alibaba: job vibe 7, 458 job voices
- #4 [Claude Sonnet 5](https://notanotherbench.ani.computer/models/anthropic-claude-sonnet-5) by Anthropic: job vibe 6.8, 3,247 job voices
- #5 [GPT 6 Astra](https://notanotherbench.ani.computer/models/openai-gpt-6-astra) by OpenAI: job vibe 6.8, 689 job voices
- #6 [DeepSeek V4 Flash](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-flash-0731) by DeepSeek: job vibe 6.8, 308 job voices
- #7 [GLM 5.3 Flash](https://notanotherbench.ani.computer/models/zhipu-glm-5-3-flash) by Zhipu: job vibe 6.7, 205 job voices
- #8 [Step 2](https://notanotherbench.ani.computer/models/stepfun-global-step-2-16k) by StepFun: job vibe 6.7, 159 job voices

## The People's Leaderboard (top community-reviewed models)

- [Grok 4.6](https://notanotherbench.ani.computer/models/xai-grok-4-6): vibe 7.7, 69,744 voices, by xAI (Multimodal, #1 in Multimodal). People love: People feel it is worth the money
- [Qwen3.8 27B](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-27b): vibe 7.6, 19,790 voices, by Alibaba (Multimodal, #2 in Multimodal). People love: Popular for coding help
- [Mercury 2.5](https://notanotherbench.ani.computer/models/inception-mercury-2-5): vibe 7.5, 6,901 voices, by Inception (Text, #1 in Text). People love: Popular for coding help
- [Fugu Max](https://notanotherbench.ani.computer/models/sakana-fugu-max): vibe 7.5, 5,959 voices, by Sakana (Multimodal, #3 in Multimodal). People love: Good at writing and wording things
- [Schematron V2 Turbo](https://notanotherbench.ani.computer/models/inference-net-schematron-v2-turbo): vibe 7.5, 5,871 voices, by Inference Net (Text, #2 in Text). People love: Usually gives correct answers
- [Muse Spark 1.3](https://notanotherbench.ani.computer/models/meta-muse-spark-1-3): vibe 7.4, 10,059 voices, by Meta (Text, #3 in Text). People love: Popular for coding help
- [V1.1](https://notanotherbench.ani.computer/models/black-forest-labs-v1-1): vibe 7.4, 7,432 voices, by Black Forest Labs (Images, #1 in Images). People love: People feel it is worth the money
- [Qwen3.8 Flash](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-flash): vibe 7.4, 7,372 voices, by Alibaba (Multimodal, #4 in Multimodal). People love: Popular for coding help
- [MiniMax Music 3](https://notanotherbench.ani.computer/models/minimax-minimax-music-3): vibe 7.4, 4,663 voices, by MiniMax (Audio, #1 in Audio). People love: Stable and dependable
- [Pixel Art Style](https://notanotherbench.ani.computer/models/kohbanye-pixel-art-style): vibe 7.4, 3,799 voices, by Kohbanye (Images, #2 in Images). People love: People feel it is worth the money
- [Elegance](https://notanotherbench.ani.computer/models/ndimensional-elegance): vibe 7.4, 3,786 voices, by Ndimensional (Video, #1 in Video). People love: Results look great
- [Dots3 Note](https://notanotherbench.ani.computer/models/xiaohongshu-dots3-note): vibe 7.4, 2,823 voices, by Xiaohongshu (Multimodal, #5 in Multimodal). People love: Popular for coding help
- [Flux](https://notanotherbench.ani.computer/models/black-forest-labs-flux): vibe 7.3, 29,832 voices, by Black Forest Labs (Images, #3 in Images). People love: Results look great
- [DeepSeek V4 Flash](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-flash-0731): vibe 7.3, 12,076 voices, by DeepSeek (Text, #4 in Text). People love: People feel it is worth the money
- [DeepSeek V4 Flash Vision Exp](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-flash-vision-exp): vibe 7.3, 5,209 voices, by DeepSeek (Multimodal, #6 in Multimodal). People love: People feel it is worth the money
- [Apertus v1.5 70B](https://notanotherbench.ani.computer/models/infomaniak-apertus-v1-5-70b): vibe 7.3, 2,532 voices, by Infomaniak (Text, #5 in Text)
- [Gemini 3.7 Flash](https://notanotherbench.ani.computer/models/google-gemini-3-7-flash): vibe 7.2, 9,727 voices, by Google (Multimodal, #7 in Multimodal). People love: Popular for coding help
- [Uncensored](https://notanotherbench.ani.computer/models/cognitivecomputations-uncensored): vibe 7.2, 3,784 voices, by Cognitivecomputations (Text, #6 in Text). People love: Popular for coding help
- [Ling 3.0 Tiny (free)](https://notanotherbench.ani.computer/models/inclusionai-ling-3-0-tiny-free): vibe 7.2, 2,313 voices, by inclusionAI (Text, #7 in Text). People love: People feel it is worth the money
- [Midjourney](https://notanotherbench.ani.computer/models/midjourney): vibe 7.1, 16,090 voices, by Midjourney (Images, #4 in Images). People love: Results look great
- [Grok Imagine](https://notanotherbench.ani.computer/models/xai-grok-imagine): vibe 7.1, 5,150 voices, by xAI (Video, #2 in Video). People love: Results look great
- [v1.1 Ultra](https://notanotherbench.ani.computer/models/black-forest-labs-v1-1-ultra): vibe 7.1, 3,299 voices, by Black Forest Labs (Images, #5 in Images). People love: People feel it is worth the money
- [Step 2](https://notanotherbench.ani.computer/models/stepfun-global-step-2-16k): vibe 7.1, 2,260 voices, by StepFun (Text, #8 in Text). People love: Popular for coding help
- [Qwen3.8 2.4T A95B](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-2-4t-a95b): vibe 7.1, 1,939 voices, by Alibaba (Multimodal, #8 in Multimodal). People love: Popular for coding help
- [Colorful](https://notanotherbench.ani.computer/models/recoilme-colorful): vibe 7.1, 1,110 voices, by Recoilme (Images, #6 in Images). People love: Results look great

## Top models by category

### Text

- #1 [Mercury 2.5](https://notanotherbench.ani.computer/models/inception-mercury-2-5) by Inception: vibe 7.5, 6,901 voices
- #2 [Schematron V2 Turbo](https://notanotherbench.ani.computer/models/inference-net-schematron-v2-turbo) by Inference Net: vibe 7.5, 5,871 voices
- #3 [Muse Spark 1.3](https://notanotherbench.ani.computer/models/meta-muse-spark-1-3) by Meta: vibe 7.4, 10,059 voices
- #4 [DeepSeek V4 Flash](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-flash-0731) by DeepSeek: vibe 7.3, 12,076 voices
- #5 [Apertus v1.5 70B](https://notanotherbench.ani.computer/models/infomaniak-apertus-v1-5-70b) by Infomaniak: vibe 7.3, 2,532 voices
- #6 [Uncensored](https://notanotherbench.ani.computer/models/cognitivecomputations-uncensored) by Cognitivecomputations: vibe 7.2, 3,784 voices
- #7 [Ling 3.0 Tiny (free)](https://notanotherbench.ani.computer/models/inclusionai-ling-3-0-tiny-free) by inclusionAI: vibe 7.2, 2,313 voices
- #8 [Step 2](https://notanotherbench.ani.computer/models/stepfun-global-step-2-16k) by StepFun: vibe 7.1, 2,260 voices

### Code

- #1 [Qwen3 Coder 480B A35B](https://notanotherbench.ani.computer/models/alibaba-qwen3-coder-480b-a35b) by Alibaba: vibe 6.6, 930 voices
- #2 [Grok Code](https://notanotherbench.ani.computer/models/xai-grok-code-fast) by xAI: vibe 6.3, 646 voices
- #3 [Qwen3 Coder Next](https://notanotherbench.ani.computer/models/alibaba-qwen3-coder-next) by Alibaba: vibe 6.2, 239 voices
- #4 [Deepseek Coder](https://notanotherbench.ani.computer/models/deepseek-deepseek-coder) by DeepSeek: vibe 6, 293 voices
- #5 [Codestral](https://notanotherbench.ani.computer/models/mistral-codestral) by Mistral: vibe 6, 224 voices
- #6 [Devstral 2](https://notanotherbench.ani.computer/models/mistral-devstral-2) by Mistral: vibe 5.8, 98 voices
- #7 [R1 Distill Llama 70B](https://notanotherbench.ani.computer/models/meta-r1-distill-llama-70b) by Meta: vibe 5.8, 77 voices
- #8 [Qwen Coder Plus (Alibaba Cloud)](https://notanotherbench.ani.computer/models/alibaba-qwen-coder-plus) by Alibaba: vibe 5.6, 88 voices

### Images

- #1 [V1.1](https://notanotherbench.ani.computer/models/black-forest-labs-v1-1) by Black Forest Labs: vibe 7.4, 7,432 voices
- #2 [Pixel Art Style](https://notanotherbench.ani.computer/models/kohbanye-pixel-art-style) by Kohbanye: vibe 7.4, 3,799 voices
- #3 [Flux](https://notanotherbench.ani.computer/models/black-forest-labs-flux) by Black Forest Labs: vibe 7.3, 29,832 voices
- #4 [Midjourney](https://notanotherbench.ani.computer/models/midjourney) by Midjourney: vibe 7.1, 16,090 voices
- #5 [v1.1 Ultra](https://notanotherbench.ani.computer/models/black-forest-labs-v1-1-ultra) by Black Forest Labs: vibe 7.1, 3,299 voices
- #6 [Colorful](https://notanotherbench.ani.computer/models/recoilme-colorful) by Recoilme: vibe 7.1, 1,110 voices
- #7 [Oil painting](https://notanotherbench.ani.computer/models/mushroom-oil-painting) by Mushroom: vibe 7, 5,256 voices
- #8 [Flux.2 Klein](https://notanotherbench.ani.computer/models/black-forest-labs-flux-2-klein) by Black Forest Labs: vibe 6.9, 5,399 voices

### Video

- #1 [Elegance](https://notanotherbench.ani.computer/models/ndimensional-elegance) by Ndimensional: vibe 7.4, 3,786 voices
- #2 [Grok Imagine](https://notanotherbench.ani.computer/models/xai-grok-imagine) by xAI: vibe 7.1, 5,150 voices
- #3 [Seedance 2.0](https://notanotherbench.ani.computer/models/bytedance-seedance-2-0) by ByteDance: vibe 6.9, 737 voices
- #4 [Runway Gen 4 Turbo](https://notanotherbench.ani.computer/models/runway-gen-4-turbo) by Runway: vibe 6.8, 894 voices
- #5 [Seedance 2.5](https://notanotherbench.ani.computer/models/bytedance-seedance-2-5) by ByteDance: vibe 6.8, 861 voices
- #6 [LTX 2](https://notanotherbench.ani.computer/models/lightricks-ltx-2) by Lightricks: vibe 6.7, 892 voices
- #7 [LTX 2.3](https://notanotherbench.ani.computer/models/zachyuan-ltx-2-3) by Lightricks: vibe 6.6, 1,149 voices
- #8 [DreamShaper](https://notanotherbench.ani.computer/models/lykon-dreamshaper) by Lykon: vibe 6.6, 393 voices

### Audio

- #1 [MiniMax Music 3](https://notanotherbench.ani.computer/models/minimax-minimax-music-3) by MiniMax: vibe 7.4, 4,663 voices
- #2 [Kokoro TTS](https://notanotherbench.ani.computer/models/kokoro-kokoro-tts) by Kokoro: vibe 7, 3,140 voices
- #3 [ElevenLabs Music](https://notanotherbench.ani.computer/models/elevenlabs-music) by ElevenLabs: vibe 6.9, 1,852 voices
- #4 [Lyria 3 Pro](https://notanotherbench.ani.computer/models/lyria3-lyria-3-pro) by Lyria3: vibe 6.2, 2,293 voices
- #5 [Gemini TTS](https://notanotherbench.ani.computer/models/google-gemini-tts) by Google: vibe 6.2, 332 voices
- #6 [Eleven](https://notanotherbench.ani.computer/models/elevenlabs-eleven) by ElevenLabs: vibe 6.2, 152 voices
- #7 [nemotron 3.5 asr streaming 0.6B](https://notanotherbench.ani.computer/models/onnx-community-nemotron-3-5-asr-streaming-0-6b-onnx-int4) by Nvidia: vibe 6.1, 463 voices
- #8 [Kokoro TTS](https://notanotherbench.ani.computer/models/kokoro-tts) by Hexgrad: vibe 6.1, 158 voices

### Multimodal

- #1 [Grok 4.6](https://notanotherbench.ani.computer/models/xai-grok-4-6) by xAI: vibe 7.7, 69,744 voices
- #2 [Qwen3.8 27B](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-27b) by Alibaba: vibe 7.6, 19,790 voices
- #3 [Fugu Max](https://notanotherbench.ani.computer/models/sakana-fugu-max) by Sakana: vibe 7.5, 5,959 voices
- #4 [Qwen3.8 Flash](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-flash) by Alibaba: vibe 7.4, 7,372 voices
- #5 [Dots3 Note](https://notanotherbench.ani.computer/models/xiaohongshu-dots3-note) by Xiaohongshu: vibe 7.4, 2,823 voices
- #6 [DeepSeek V4 Flash Vision Exp](https://notanotherbench.ani.computer/models/deepseek-deepseek-v4-flash-vision-exp) by DeepSeek: vibe 7.3, 5,209 voices
- #7 [Gemini 3.7 Flash](https://notanotherbench.ani.computer/models/google-gemini-3-7-flash) by Google: vibe 7.2, 9,727 voices
- #8 [Qwen3.8 2.4T A95B](https://notanotherbench.ani.computer/models/alibaba-qwen3-8-2-4t-a95b) by Alibaba: vibe 7.1, 1,939 voices

## Key pages

- [Home / full catalog and leaderboard](https://notanotherbench.ani.computer)
- [Best models for writing](https://notanotherbench.ani.computer/leaderboard/writing)
- [Best models for coding](https://notanotherbench.ani.computer/leaderboard/coding)
- [Best models for agents](https://notanotherbench.ani.computer/leaderboard/agents)
- [Best models for research](https://notanotherbench.ani.computer/leaderboard/research)
- [Compare models](https://notanotherbench.ani.computer/compare)
- [How the board works](https://notanotherbench.ani.computer/methodology)
- [Open stats](https://notanotherbench.ani.computer/stats)
- [Sitemap](https://notanotherbench.ani.computer/sitemap.xml)

Model pages live at /models/<slug> and include machine-readable JSON-LD structured data.