# Frontend and web development leaderboard

As of November 21, 2026, Claude Fable 5.2 leads our frontend and web development leaderboard at 89.5 (Supported, $8/$40 per 1M tokens), ahead of Claude Opus 5.5 at 82.8. GPT-6.1 Sol (#3, 81.2) is the lowest-cost model near the top at $2/$10 per 1M tokens; Kimi K3.1 Pro leads open-weight models at #15 (70.0). We rank 76 models (38 Supported, 38 Estimated) across 9 domain protocols, and this vertical carries zero weight in the overall score.

## Shortlist by workload and budget

- **Highest point score:** [Claude Fable 5.2](/models/claude-fable-5-2) (#1, 89.5, Supported, $8/$40 per 1M tokens) — Costs $8/$40 per 1M tokens; GPT-6.1 Sol (#3, 81.2) cuts token spend to $2/$10 per 1M tokens.
- **Closest contender:** [Claude Opus 5.5](/models/claude-opus-5-5) (#2, 82.8, Supported, $4/$20 per 1M tokens) — Sits 6.7 points behind Claude Fable 5.2; test both on your own prompts when provider lock-in or latency matters.
- **Best API value:** [GPT-6.1 Sol](/models/gpt-6-1-sol) (#3, 81.2, Supported, $2/$10 per 1M tokens) — Keeps 91% of the top score (81.2 vs 89.5), but gives up 8.3 points on the hardest tail cases.
- **Top open-weight model:** [Kimi K3.1 Pro](/models/kimi-3-1-pro) (#15, 70.0, Supported, Open weights · $2/$10 per 1M tokens) — Trails Claude Fable 5.2 by 19.5 points (#15 overall here), so pick it when weight control or on-prem residency outweighs peak score.

## Ranked models

| Rank | Model | Creator | Score | Conditional range | Evidence | API ($/1M in/out) | Context | Source families |
|---:|---|---|---:|---:|---|---:|---:|---:|
| #1 | [Claude Fable 5.2](/models/claude-fable-5-2) | Anthropic | 89.5 | 75.9-100.0 | Supported | $8.00 / $40.00 | 1M | 2 |
| #2 | [Claude Opus 5.5](/models/claude-opus-5-5) | Anthropic | 82.8 | 72.9-92.7 | Supported | $4.00 / $20.00 | 1M | 6 |
| #3 | [GPT-6.1 Sol](/models/gpt-6-1-sol) | OpenAI | 81.2 | 71.0-91.4 | Supported | $2.00 / $10.00 | 1.05M | 4 |
| #4 | [GPT-6 Astra](/models/gpt-6-astra) | OpenAI | 80.2 | 70.2-90.2 | Supported | $10.00 / $50.00 | 1.05M | 6 |
| #5 | [Claude Fable 5.1](/models/claude-fable-5-1) | Anthropic | 77.6 | 66.2-89.0 | Supported | $10.00 / $50.00 | 1M | 3 |
| #6 | [Claude Sonnet 5.5](/models/claude-sonnet-5-5) | Anthropic | 78.0 | 65.7-90.3 | Estimated | $2.00 / $10.00 | 1M | 1 |
| #7 | [Gemini 4 Carbon](/models/gemini-4-carbon) | Google | 77.9 | 65.5-90.3 | Estimated | $2.00 / $10.00 | 1M | 1 |
| #8 | [GPT-6 Sol](/models/gpt-6-sol) | OpenAI | 75.8 | 65.2-86.4 | Supported | $2.00 / $10.00 | 1.05M | 2 |
| #9 | [GPT-6.1 Terra](/models/gpt-6-1-terra) | OpenAI | 75.8 | 65.4-86.2 | Supported | $1.00 / $5.00 | 1.05M | 2 |
| #10 | [Claude Opus 5](/models/claude-opus-5) | Anthropic | 74.3 | 64.1-84.5 | Supported | $5.00 / $25.00 | - | 3 |
| #11 | [GPT-5.6 Sol](/models/gpt-5-6-sol) | OpenAI | 74.0 | 63.9-84.1 | Supported | $4.00 / $20.00 | 1.05M | 4 |
| #12 | [Qwen4 Max](/models/qwen4-max) | Alibaba | 74.0 | 60.1-87.8 | Estimated | $1.50 / $4.50 | 1M | 1 |
| #13 | [Gemini 4 Argon](/models/gemini-4-argon) | Google | 71.9 | 61.8-82.0 | Supported | $2.00 / $10.00 | 1M | 5 |
| #14 | [Claude Fable 5](/models/claude-fable) | Anthropic | 70.7 | 60.3-81.1 | Supported | $10.00 / $50.00 | 1M+ | 2 |
| #15 | [Kimi K3.1 Pro](/models/kimi-3-1-pro) | Moonshot AI | 70.0 | 58.9-81.0 | Supported | $2.00 / $10.00 | 1.05M | 2 |
| #16 | [Gemini 4 Hydrogen](/models/gemini-4-hydrogen) | Google | 71.6 | 59.3-84.0 | Estimated | $1.50 / $7.50 | 1M | 1 |
| #17 | [DeepSeek V4.1 Flash 1017](/models/deepseek-v4-1-flash) | DeepSeek | 69.2 | 58.9-79.4 | Supported | $0.30 / $1.20 | 1M | 3 |
| #18 | [Grok 4.6](/models/grok-4-6) | xAI | 69.1 | 59.1-79.2 | Supported | $2.00 / $6.00 | 500K | 4 |
| #19 | [GPT-5.5](/models/gpt-5-5) | OpenAI | 66.5 | 56.4-76.6 | Supported | $5.00 / $30.00 | 1M | 4 |
| #20 | [Claude Opus 4.8](/models/claude-opus-4-8) | Anthropic | 66.2 | 54.2-78.2 | Estimated | $5.00 / $25.00 | 1M | 2 |
| #21 | [GPT-5.6 Terra](/models/gpt-5-6-terra) | OpenAI | 65.9 | 55.6-76.2 | Supported | $2.00 / $12.00 | 1.05M | 2 |
| #22 | [Qwen3.8 Max](/models/qwen3-8-max) | Alibaba | 65.8 | 55.5-76.1 | Supported | - | 1M | 3 |
| #23 | [GPT-6 Luna](/models/gpt-6-luna) | OpenAI | 64.9 | 54.6-75.1 | Supported | $0.10 / $0.50 | 1.05M | 3 |
| #24 | [GPT-5.5 Pro](/models/gpt-5-5-pro) | OpenAI | 66.7 | 53.6-79.8 | Estimated | $30.00 / $180.00 | 1M | 1 |
| #25 | [Gemini 3.8 Flash](/models/gemini-3-8-flash) | Google | 64.6 | 54.6-74.6 | Supported | $0.75 / $3.75 | 1M | 6 |
| #26 | [Kimi K3](/models/kimi-k3) | Moonshot AI | 64.6 | 53.6-75.6 | Supported | $3.00 / $15.00 | 1.05M | 5 |
| #27 | [MiniMax M3.1 Flash](/models/minimax-m3-1-flash) | MiniMax | 64.3 | 51.0-77.6 | Supported | $0.15 / $0.60 | 1M | 2 |
| #28 | [Gemini 3.7 Flash](/models/gemini-3-7-flash) | Google | 64.3 | 53.8-74.9 | Supported | $0.75 / $3.75 | 1M | 2 |
| #29 | [GPT-5.6 Luna](/models/gpt-5-6-luna) | OpenAI | 63.9 | 53.3-74.4 | Supported | $0.20 / $1.20 | 1.05M | 2 |
| #30 | [MiMo-V2.6-Pro](/models/mimo-v2-6-pro) | Xiaomi | 65.7 | 50.9-80.4 | Estimated | $0.43 / $0.87 | 1M | 1 |
| #31 | [Claude Haiku 5.5](/models/claude-haiku-5-5) | Anthropic | 63.6 | 53.1-74.1 | Supported | $0.10 / $0.50 | 1M | 2 |
| #32 | [GLM-5.3](/models/glm-5-3) | Z.AI | 63.5 | 53.2-73.7 | Supported | Free | 1M | 3 |
| #33 | [Muse Wonder](/models/muse-wonder) | Meta | 65.4 | 54.2-76.6 | Estimated | $4.00 / $20.00 | 1M | 1 |
| #34 | [GLM-5.3-Flash](/models/glm-5-3-flash) | Z.AI | 63.2 | 51.7-74.7 | Supported | Free | 1M | 3 |
| #35 | [Gemini 3.6 Flash](/models/gemini-3-6-flash) | Google | 64.1 | 53.4-74.8 | Estimated | $0.75 / $3.75 | 1M | 1 |
| #36 | [Qwen3.8-27B](/models/qwen3-8-27b) | Alibaba | 62.4 | 49.4-75.4 | Estimated | Free | 262K | 1 |
| #37 | [Gemini 3.1 Pro](/models/gemini-3-1-pro) | Google | 60.2 | 50.0-70.3 | Supported | $2.00 / $12.00 | 1M | 4 |
| #38 | [Grok 4.5](/models/grok-4-5) | xAI | 62.1 | 51.4-72.8 | Estimated | $2.00 / $6.00 | 500K | 1 |
| #39 | [Step 5 Preview](/models/step-5-preview) | StepFun | 61.7 | 46.5-76.9 | Estimated | $1.00 / $2.70 | 1M | 1 |
| #40 | [GLM-5.2](/models/glm-5-2) | Z.AI | 58.8 | 48.5-69.1 | Supported | $1.40 / $4.40 | 1M | 3 |
| #41 | [GPT-5.4](/models/gpt-5-4) | OpenAI | 58.4 | 48.0-68.8 | Supported | $2.50 / $15.00 | 1.05M | 4 |
| #42 | [Gemini 3.5 Flash](/models/gemini-3-5-flash) | Google | 60.4 | 49.7-71.0 | Estimated | $1.50 / $9.00 | 1M | 1 |
| #43 | [DeepSeek V4 Pro](/models/deepseek-v4-pro) | DeepSeek | 60.0 | 49.2-70.7 | Estimated | $1.32 / $3.96 | 1M | 1 |
| #44 | [Muse Spark 1.2](/models/muse-spark-1-2) | Meta | 57.8 | 43.0-72.7 | Supported | $1.25 / $4.25 | 1M | 2 |
| #45 | [Claude Opus 4.7](/models/claude-opus-4-7) | Anthropic | 57.4 | 46.1-68.7 | Supported | $5.00 / $25.00 | 1M | 2 |
| #46 | [Claude Sonnet 5](/models/claude-sonnet-5) | Anthropic | 57.2 | 39.0-75.4 | Supported | $2.00 / $10.00 | 1M | 2 |
| #47 | [Claude Opus 4.6](/models/claude-opus-4-6) | Anthropic | 57.2 | 46.1-68.3 | Supported | $5.00 / $25.00 | 1M | 3 |
| #48 | [Claude Opus 4.6 (Adaptive)](/models/claude-opus-4-6-thinking) | Anthropic | 57.2 | 46.1-68.3 | Supported | - | 1M | 3 |
| #49 | [Qwen3.7 Max](/models/qwen3-7-max) | Alibaba | 58.3 | 47.7-68.9 | Estimated | - | 1M | 1 |
| #50 | [Muse Spark 1.1](/models/muse-spark-1-1) | Meta | 56.8 | 45.7-67.9 | Estimated | - | 1M | 1 |
| #51 | [Gemini 3 Pro](/models/gemini-3-pro) | Google | 54.0 | 43.1-65.0 | Supported | $2.00 / $12.00 | 2M | 3 |
| #52 | [GLM-5.1](/models/glm-5-1) | Z.AI | 54.5 | 42.9-66.1 | Estimated | $1.40 / $4.40 | 203K | 1 |
| #53 | [Claude Opus 4.5](/models/claude-opus-4-5) | Anthropic | 52.3 | 41.6-63.1 | Estimated | $5.00 / $25.00 | 200K | 2 |
| #54 | [Claude Opus 4.5 Thinking](/models/claude-opus-4-5-thinking) | Anthropic | 52.3 | 41.6-63.1 | Estimated | - | 200K | 2 |
| #55 | [MiniMax M3](/models/minimax-m3) | MiniMax | 52.1 | 37.3-67.0 | Supported | $0.30 / $1.20 | 1M | 2 |
| #56 | [Gemini 3 Flash](/models/gemini-3-flash) | Google | 53.8 | 38.4-69.2 | Estimated | $0.50 / $3.00 | 1M | 1 |
| #57 | [Claude Sonnet 4.6](/models/claude-sonnet-4-6) | Anthropic | 51.9 | 35.1-68.8 | Estimated | $3.00 / $15.00 | 200K | 1 |
| #58 | [GPT-5.4 mini](/models/gpt-5-4-mini) | OpenAI | 51.4 | 36.7-66.1 | Estimated | $0.75 / $4.50 | 400K | 1 |
| #59 | [GPT-5](/models/gpt-5) | OpenAI | 49.5 | 38.0-61.0 | Estimated | $1.25 / $10.00 | 128K | 1 |
| #60 | [Mistral Large 4](/models/mistral-large-4) | Mistral | 48.9 | 35.5-62.2 | Estimated | $0.70 / $2.10 | 1M | 1 |
| #61 | [Qwen3.5 397B](/models/qwen3-5-397b) | Alibaba | 46.0 | 32.7-59.3 | Estimated | $0.60 / $3.60 | 128K | 2 |
| #62 | [Claude Sonnet 4.5](/models/claude-sonnet-4-5) | Anthropic | 46.2 | 31.4-61.0 | Estimated | $3.00 / $15.00 | 200K | 1 |
| #63 | [Claude Sonnet 4.5 Thinking](/models/claude-sonnet-4-5-thinking) | Anthropic | 46.2 | 31.4-61.0 | Estimated | - | 200K | 1 |
| #64 | [Muse Spark Open](/models/muse-spark-open) | Meta | 44.2 | 29.2-59.3 | Estimated | - | 1M | 1 |
| #65 | [Grok 4](/models/grok-4) | xAI | 43.1 | 24.6-61.7 | Estimated | - | 128K | 1 |
| #66 | [Kimi K2.5](/models/kimi-k2-5) | Moonshot AI | 41.1 | 25.8-56.5 | Supported | $0.60 / $3.00 | 256K | 2 |
| #67 | [Gemma 4 31B](/models/gemma-4-31b) | Google | 40.6 | 26.7-54.4 | Supported | Free | 256K | 2 |
| #68 | [GLM-5](/models/glm-5) | Z.AI | 41.6 | 24.5-58.8 | Estimated | $1.00 / $3.20 | 200K | 1 |
| #69 | [Gemini 2.5 Pro](/models/gemini-2-5-pro) | Google | 40.4 | 28.1-52.7 | Estimated | $1.25 / $10.00 | 1M | 1 |
| #70 | [Inkling](/models/inkling) | Thinking Machines Lab | 37.5 | 25.1-49.9 | Estimated | $1.87 / $4.68 | 1M | 1 |
| #71 | [MiniMax M2.7](/models/minimax-m2-7) | MiniMax | 36.2 | 17.8-54.6 | Estimated | $0.30 / $1.20 | 200K | 1 |
| #72 | [DeepSeek V3.2](/models/deepseek-v3-2) | DeepSeek | 34.7 | 16.2-53.2 | Estimated | $0.28 / $0.42 | 128K | 1 |
| #73 | [GPT-OSS 120B](/models/gpt-oss-120b) | OpenAI | 33.8 | 16.6-51.0 | Estimated | Free | 128K | 1 |
| #74 | [Claude Haiku 4.5](/models/claude-haiku-4-5) | Anthropic | 31.9 | 16.3-47.5 | Estimated | $1.00 / $5.00 | 200K | 1 |
| #75 | [Claude Haiku 4.5 Thinking](/models/claude-haiku-4-5-thinking) | Anthropic | 31.9 | 16.3-47.5 | Estimated | - | 200K | 1 |
| #76 | [GPT-OSS 20B](/models/gpt-oss-20b) | OpenAI | 29.8 | 11.7-47.9 | Estimated | Free | 128K | 1 |

## Scored domain benchmarks (9)

- [ScreenSpot Pro](/benchmarks/screenspot-pro) (17% ref. weight): A GUI-grounding benchmark for 1,581 instructions in full-screen, high-resolution professional interfaces. It tests where a target is, not whether an agent can finish the surrounding workflow.
- [VoxelBench Text](/benchmarks/voxelbench) (16% ref. weight): A live human-preference benchmark where language models turn text prompts into voxel structures and voters compare anonymous builds from the same prompt.
- [LiveBench 2.0 Design](/benchmarks/livebench) · LiveBench (16% ref. weight): Contamination-resistant frontend interface and interactive component design tasks.
- [GameDevBench](/benchmarks/gamedevbench) (14% ref. weight): Evaluates coding agents on 333 multimodal game-development tasks in Godot, spanning 2D graphics, 3D graphics, user interfaces, and gameplay logic.
- [ReactBench](/benchmarks/reactbench) (11% ref. weight): A coding-agent benchmark for realistic React work, with rubrics that check production concerns such as performance, accessibility, correctness, and code quality.
- [Code2Video](/benchmarks/code2video) (11% ref. weight): HeyGen's Code2Video tests whether a model can turn a detailed creative brief into a finished motion-graphics video by writing code.
- [Web Clone Bench](/benchmarks/web-clone-bench) (8% ref. weight): Coding agents recreate nine reference web apps from an empty directory. Plato checks visual and functional behavior through end-to-end workflows.
- [React Native Evals](/benchmarks/reactnativeevals) (7% ref. weight): An open benchmark for AI coding agents on real-world React Native implementation tasks, emphasizing working app behavior, recommended architecture choices, and strict constraint adherence.
- [Next.js Evals](/benchmarks/nextjsevals) (2% ref. weight): A Vercel benchmark for AI coding agents on Next.js code generation and migration tasks, reporting pass@4, average execution time, and results with bundled documentation through AGENTS.md.

## Executable React components, visual layout cloning, and pixel-level UI grounding

Nine benchmarks shape this vertical, led by ScreenSpot Pro (17%), VoxelBench Text (16%), LiveBench 2.0 Design (16%) alongside GameDevBench, ReactBench, Code2Video, Web Clone Bench, React Native Evals, and Next.js Evals. Rather than grading isolated Python algorithms, these suites execute real component trees, check interactive state updates in React and Next.js, ground cursor coordinates on high-density desktop viewports, and render 2D/3D browser scenes.

Thirty-eight of the 76 models hold Supported status because multiple independent evaluators—including ScreenSpot Pro, LiveBench, and framework-specific boards—test the same public checkpoints.

## Passing component assertions does not guarantee tasteful UI or accessible markup

Automated DOM harnesses and visual-similarity metrics reward layout match and state transitions, but they rarely catch generic SaaS card grids, awkward mobile tap targets, missing ARIA focus traps, or bloated client bundles. A model can ace ReactBench while still importing deprecated Next.js router patterns unless your prompt supplies current framework docs.

Keep Lighthouse, axe-core accessibility checks, and bundle-size budgets in continuous integration regardless of which model writes your JSX.

[Frontend & app-dev task guide](/best/frontend-app-dev) · [Claude Fable 5.2 vs Claude Opus 5.5](/compare/claude-fable-5-2-vs-claude-opus-5-5)

## When to pay for the frontier leader versus fast mid-tier iteration

Claude Fable 5.2 (89.5, Supported, $8/$40 per 1M tokens) holds a 6.7-point lead over Claude Opus 5.5 (82.8, Supported, $4/$20 per 1M tokens), one of the widest #1-to-#2 gaps on any BenchAlign v6.0 board. That margin shows up on complex multi-file React refactors and interactive canvas tasks where a single broken hook stalls the preview.

For high-frequency component scaffolding or design-system variants where $8/$40 tokens add up quickly, GPT-6.1 Sol (#3, 81.2, Supported at $2/$10 per 1M tokens) cuts API spend by 75% while staying above 81 points, and Kimi K3.1 Pro (#15, 70.0, Supported) leads self-hosted weights.

[General coding leaderboard](/coding) · [Compare AI coding subscriptions](/coding-plans)

## Questions

### Which LLM is best for frontend development right now?

Claude Fable 5.2 holds #1 on our frontend development leaderboard at 89.5 (Supported) as of November 21, 2026, ahead of Claude Opus 5.5 at 82.8 and GPT-6.1 Sol at 81.2. Both top models hold Supported status across independent public evaluations.

### What is the best open-weight LLM for frontend development?

Kimi K3.1 Pro from Moonshot AI leads open-weight models on this frontend development board at #15 with a calibrated score of 70.0 (Supported). That trails Claude Fable 5.2 by 19.5 points, so choose Kimi K3.1 Pro when self-hosting, VPC data residency, or weight customization matters more than peak frontier accuracy.

### Which API model offers the best value for frontend development?

GPT-6.1 Sol (#3, 81.2) is our top API value pick for frontend development, priced at $2/$10 per 1M tokens compared with Claude Fable 5.2 at $8/$40 per 1M tokens. You give up 8.3 calibrated points, making GPT-6.1 Sol the better buy whenever those extra points do not cut downstream retries.

### Which benchmarks make up the frontend development leaderboard score?

BenchAlign v6.0 calibrates 9 open and primary domain protocols for this board, led by ScreenSpot Pro (17%), VoxelBench Text (16%), LiveBench 2.0 Design (16%), GameDevBench (14%). Each percentage reflects share of the vertical's reference weight inside our item-response model rather than a raw score average, and no single independent evaluator can exceed 40% when three or more evaluators contribute.

### Does the frontend leaderboard test real React and Next.js execution?

Yes. ReactBench, Next.js Evals, React Native Evals, Web Clone Bench, and GameDevBench execute generated component code and browser apps against functional assertions and visual layout targets, while ScreenSpot Pro tests high-resolution UI coordinate grounding. Static syntax fluency alone cannot earn a top frontend score here.

### What frontend qualities are missing from automated web-dev benchmarks?

Automated UI benchmarks verify DOM state, layout geometry, and task completion, but they do not grade brand typography, keyboard focus management, screen-reader WCAG compliance, Core Web Vitals bundle weight, or API security headers. Keep automated accessibility and bundle audits in CI alongside your chosen coding model.

## Related benchmarks and guides

- [Coding Leaderboard](/coding): Backend, repository patch, and competitive programming rankings.
- [Best LLMs for Frontend & App Dev](/best/frontend-app-dev): Task guide for React, mobile UI, and design-to-code workflows.
- [AI Coding Plans Compared](/coding-plans): Compare Claude Code, Codex, Cursor, and subscription limits.
- [Best Computer-Use Models](/best/computer-use): GUI grounding and browser-control rankings.

## Other specialist vertical leaderboards

- [Cybersecurity leaderboard](/leaderboards/cybersecurity) — #1 Gemini 4 Carbon (90.0)
- [Documents and extraction leaderboard](/leaderboards/documents) — #1 GPT-6.2 Astra (86.5)
- [Factuality and anti-hallucination leaderboard](/leaderboards/factuality) — #1 Gemini 4 Carbon (78.8)
- [Finance and accounting leaderboard](/leaderboards/finance) — #1 Gemini 4 Carbon (89.2)
- [Healthcare and medical leaderboard](/leaderboards/healthcare) — #1 Claude Fable 5.2 (87.5)
- [Legal leaderboard](/leaderboards/legal) — #1 Claude Fable 5.2 (86.1)
- [Science and research leaderboard](/leaderboards/science) — #1 Claude Fable 5.2 (94.1)
- [Search and deep research leaderboard](/leaderboards/search) — #1 Claude Fable 5.2 (86.6)

Last updated: November 21, 2026

Canonical page: https://benchlm.ai/leaderboards/frontend
