Benchmark Almanac

React Native Evals

Claude Fable 5.1 89% leads since September 2026
Not in model cards

Models implement React Native features from task descriptions and fixed project contexts. A model judge checks the submitted files against explicit requirements, and evaluation aggregates requirement scores across tasks.

The current repository contains 134 evaluations. Comparisons must retain the task revision and judge configuration.

Made by
Callstack
Released
March 2026
Kind
Web and app front ends
Level
University
Scored
Graded by a model
Answers
Code
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Code → Code
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
GPT-6 Astra #4 on the capability index
Last model added
September 2026
Scores refreshed
November 18th, 2026
32
Entries on its own leaderboard, 27 models we track

Claude Fable 5.1 leads with 89% among models released by September 2026 Claude Fable 5.1 leads with 89% among models released by September 2026

Best score per entry from the benchmark's own leaderboard, read on November 18th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

The newest model with a score here came out in September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Claude Fable 5.1 2026-10-01; 10 runs 89%
  2. 2 Claude Opus 5.5 2026-10-01; 10 runs 88.6%
  3. 3 Claude Opus 5 2026-10-01; 10 runs 88.1%
  4. 4 apex/callstack/Apex 2026-10-01; 10 runs 87.8%
  5. 5 Claude Fable 5 2026-10-01; 10 runs 87%
  6. 6 GPT-6 Astra 2026-10-01; 10 runs 86.4%
  7. 7 GPT-5.6 Sol 2026-10-01; 10 runs 86.3%
  8. 8 GPT-6.1 Sol 2026-10-01; 10 runs 86.3%
  9. 9 Claude Opus 4.8 2026-10-01; 10 runs 84.7%
  10. 10 GPT-5.5 2026-10-01; 10 runs 84.7%
  11. 11 Grok 4.5 2026-10-01; 10 runs 84.6%
  12. 12 Grok 4.6 2026-10-01; 10 runs 84.1%
  13. 13 GPT-5.6 Terra 2026-10-01; 10 runs 83.5%
  14. 14 Claude Sonnet 5 2026-10-01; 10 runs 83.4%
  15. 15 GPT-6 Sol 2026-10-01; 10 runs 83.3%
  16. 16 GPT-5.6 Luna 2026-10-01; 10 runs 82.9%
  17. 17 openrouter/stealth/ox-alpha 2026-10-01; 10 runs 81.4%
  18. 18 Muse Spark 1.1 2026-10-01; 10 runs 81.1%
  19. 19 Gemini 3.1 Pro 2026-10-01; 10 runs 78.9%
  20. 20 MiniMax-M3 2026-10-01; 10 runs 78.6%
  21. 21 GLM-5.2 2026-10-01; 10 runs 78.3%
  22. 22 Kimi K2.6 2026-10-01; 10 runs 77.7%
  23. 23 Gemini 3.5 Flash 2026-10-01; 10 runs 75.5%
  24. 24 llama.cpp/google/gemma-4-31B-it 2026-10-01; 10 runs 75.2%
  25. 25 MiMo-V2.5-Pro 2026-10-01; 10 runs 73.4%
  26. 26 DeepSeek-V4-Flash 2026-10-01; 10 runs 73.2%
  27. 27 gpt-oss-120b 2026-10-01; 10 runs 71.6%
  28. 28 gpt-oss-20b 2026-10-01; 10 runs 71%
  29. 29 Mistral Large 3 2026-10-01; 10 runs 65.5%
  30. 30 nemotron-3-ultra-550b-a55b 2026-10-01; 10 runs 61.4%
  31. 31 nemotron-3-super-120b-a12b 2026-10-01; 10 runs 52.5%
  32. 32 Nemotron 3 Nano 30B-A3B 2026-10-01; 10 runs 42.7%

Source: Official React Native Evals leaderboard

Source: Official React Native Evals leaderboard

#EntryScoreSettingDateSource
1 Claude Fable 5.1 89% 2026-10-01; 10 runs Sep 2026 Official leaderboard
2 Claude Opus 5.5 88.6% 2026-10-01; 10 runs Sep 2026 Official leaderboard
3 Claude Opus 5 88.1% 2026-10-01; 10 runs Jul 2026 Official leaderboard
4 apex/callstack/Apex 87.8% 2026-10-01; 10 runs Official leaderboard
5 Claude Fable 5 87% 2026-10-01; 10 runs Jun 2026 Official leaderboard
6 GPT-6 Astra 86.4% 2026-10-01; 10 runs Sep 2026 Official leaderboard
7 GPT-5.6 Sol 86.3% 2026-10-01; 10 runs Jul 2026 Official leaderboard
8 GPT-6.1 Sol 86.3% 2026-10-01; 10 runs Sep 2026 Official leaderboard
9 Claude Opus 4.8 84.7% 2026-10-01; 10 runs May 2026 Official leaderboard
10 GPT-5.5 84.7% 2026-10-01; 10 runs Apr 2026 Official leaderboard
11 Grok 4.5 84.6% 2026-10-01; 10 runs Jul 2026 Official leaderboard
12 Grok 4.6 84.1% 2026-10-01; 10 runs Aug 2026 Official leaderboard
13 GPT-5.6 Terra 83.5% 2026-10-01; 10 runs Jul 2026 Official leaderboard
14 Claude Sonnet 5 83.4% 2026-10-01; 10 runs Jun 2026 Official leaderboard
15 GPT-6 Sol 83.3% 2026-10-01; 10 runs Sep 2026 Official leaderboard
16 GPT-5.6 Luna 82.9% 2026-10-01; 10 runs Jul 2026 Official leaderboard
17 openrouter/stealth/ox-alpha 81.4% 2026-10-01; 10 runs Official leaderboard
18 Muse Spark 1.1 81.1% 2026-10-01; 10 runs Jul 2026 Official leaderboard
19 Gemini 3.1 Pro 78.9% 2026-10-01; 10 runs Feb 2026 Official leaderboard
20 MiniMax-M3 78.6% 2026-10-01; 10 runs Jun 2026 Official leaderboard
21 GLM-5.2 78.3% 2026-10-01; 10 runs Jun 2026 Official leaderboard
22 Kimi K2.6 77.7% 2026-10-01; 10 runs Apr 2026 Official leaderboard
23 Gemini 3.5 Flash 75.5% 2026-10-01; 10 runs May 2026 Official leaderboard
24 llama.cpp/google/gemma-4-31B-it 75.2% 2026-10-01; 10 runs Official leaderboard
25 MiMo-V2.5-Pro 73.4% 2026-10-01; 10 runs Apr 2026 Official leaderboard
26 DeepSeek-V4-Flash 73.2% 2026-10-01; 10 runs Apr 2026 Official leaderboard
27 gpt-oss-120b 71.6% 2026-10-01; 10 runs Aug 2025 Official leaderboard
28 gpt-oss-20b 71% 2026-10-01; 10 runs Aug 2025 Official leaderboard
29 Mistral Large 3 65.5% 2026-10-01; 10 runs Dec 2025 Official leaderboard
30 nemotron-3-ultra-550b-a55b 61.4% 2026-10-01; 10 runs Official leaderboard
31 nemotron-3-super-120b-a12b 52.5% 2026-10-01; 10 runs Official leaderboard
32 Nemotron 3 Nano 30B-A3B 42.7% 2026-10-01; 10 runs Dec 2025 Official leaderboard
#ModelScoreSettingDateSource
1 Claude Fable 5.1 89% 2026-10-01; 10 runs Sep 2026 Official leaderboard
2 Claude Opus 5.5 88.6% 2026-10-01; 10 runs Sep 2026 Official leaderboard
3 Claude Opus 5 88.1% 2026-10-01; 10 runs Jul 2026 Official leaderboard
4 Claude Fable 5 87% 2026-10-01; 10 runs Jun 2026 Official leaderboard
5 GPT-6 Astra 86.4% 2026-10-01; 10 runs Sep 2026 Official leaderboard
6 GPT-5.6 Sol 86.3% 2026-10-01; 10 runs Jul 2026 Official leaderboard
7 GPT-6.1 Sol 86.3% 2026-10-01; 10 runs Sep 2026 Official leaderboard
8 Claude Opus 4.8 84.7% 2026-10-01; 10 runs May 2026 Official leaderboard
9 GPT-5.5 84.7% 2026-10-01; 10 runs Apr 2026 Official leaderboard
10 Grok 4.5 84.6% 2026-10-01; 10 runs Jul 2026 Official leaderboard
11 Grok 4.6 84.1% 2026-10-01; 10 runs Aug 2026 Official leaderboard
12 GPT-5.6 Terra 83.5% 2026-10-01; 10 runs Jul 2026 Official leaderboard
13 Claude Sonnet 5 83.4% 2026-10-01; 10 runs Jun 2026 Official leaderboard
14 GPT-6 Sol 83.3% 2026-10-01; 10 runs Sep 2026 Official leaderboard
15 GPT-5.6 Luna 82.9% 2026-10-01; 10 runs Jul 2026 Official leaderboard
16 Muse Spark 1.1 81.1% 2026-10-01; 10 runs Jul 2026 Official leaderboard
17 Gemini 3.1 Pro 78.9% 2026-10-01; 10 runs Feb 2026 Official leaderboard
18 MiniMax-M3 78.6% 2026-10-01; 10 runs Jun 2026 Official leaderboard
19 GLM-5.2 78.3% 2026-10-01; 10 runs Jun 2026 Official leaderboard
20 Kimi K2.6 77.7% 2026-10-01; 10 runs Apr 2026 Official leaderboard
21 Gemini 3.5 Flash 75.5% 2026-10-01; 10 runs May 2026 Official leaderboard
22 MiMo-V2.5-Pro 73.4% 2026-10-01; 10 runs Apr 2026 Official leaderboard
23 DeepSeek-V4-Flash 73.2% 2026-10-01; 10 runs Apr 2026 Official leaderboard
24 gpt-oss-120b 71.6% 2026-10-01; 10 runs Aug 2025 Official leaderboard
25 gpt-oss-20b 71% 2026-10-01; 10 runs Aug 2025 Official leaderboard
26 Mistral Large 3 65.5% 2026-10-01; 10 runs Dec 2025 Official leaderboard
27 Nemotron 3 Nano 30B-A3B 42.7% 2026-10-01; 10 runs Dec 2025 Official leaderboard

The best score, 89%, costs $92 for the full run

Runs on the cost frontier score higher than any run that's cheaper. GPT-6.1 Sol at 2026-10-01; 10 runs comes within 2.7 points of the best for $12 for the full run, 8 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

OpenAI MiniMax Anthropic Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.10 $1.00 $10 $100 $1,000 Cost of the full run, log scale Claude Fable 5.1, 2026-10-01; 10 runs GPT-5.6 Luna, 2026-10-01; 10 runs MiniMax-M3, 2026-10-01; 10 runs gpt-oss-120b, 2026-10-01; 10 runs 0% 20% 40% 60% 80% 100% $0.10 $1.00 $10 $100 $1,000 Cost of the full run, log scale Claude Fable 5.1, 2026-10-01; 10 runs GPT-5.6 Luna, 2026-10-01; 10 runs

Source: Official React Native Evals leaderboard

ModelEffortHarnessScoreCost of the full runOutput tokens
Claude Fable 5.1 2026-10-01; 10 runs 89% $92
Claude Opus 5.5 2026-10-01; 10 runs 88.6% $29
Claude Opus 5 2026-10-01; 10 runs 88.1% $142
Claude Fable 5 2026-10-01; 10 runs 87% $137
GPT-6 Astra 2026-10-01; 10 runs 86.4% $65
GPT-5.6 Sol 2026-10-01; 10 runs 86.3% $51
GPT-6.1 Sol 2026-10-01; 10 runs 86.3% $12
Claude Opus 4.8 2026-10-01; 10 runs 84.7% $64
GPT-5.5 2026-10-01; 10 runs 84.7% $58
Grok 4.5 2026-10-01; 10 runs 84.6% $43
Grok 4.6 2026-10-01; 10 runs 84.1% $43
GPT-5.6 Terra 2026-10-01; 10 runs 83.5% $17
Claude Sonnet 5 2026-10-01; 10 runs 83.4% $99
GPT-6 Sol 2026-10-01; 10 runs 83.3% $21
GPT-5.6 Luna 2026-10-01; 10 runs 82.9% $4.57
Muse Spark 1.1 2026-10-01; 10 runs 81.1% $19
Gemini 3.1 Pro 2026-10-01; 10 runs 78.9% $49
MiniMax-M3 2026-10-01; 10 runs 78.6% $4.44
GLM-5.2 2026-10-01; 10 runs 78.3% $92
Kimi K2.6 2026-10-01; 10 runs 77.7%
Gemini 3.5 Flash 2026-10-01; 10 runs 75.5%
MiMo-V2.5-Pro 2026-10-01; 10 runs 73.4% $6.25
DeepSeek-V4-Flash 2026-10-01; 10 runs 73.2%
gpt-oss-120b 2026-10-01; 10 runs 71.6% $1.94
gpt-oss-20b 2026-10-01; 10 runs 71% $0.79
Mistral Large 3 2026-10-01; 10 runs 65.5% $13
Nemotron 3 Nano 30B-A3B 2026-10-01; 10 runs 42.7%

The best score, 89%, costs $92 for the full run

Runs on the cost frontier score higher than any run that's cheaper. GPT-6.1 Sol at 2026-10-01; 10 runs comes within 2.7 points of the best for $12 for the full run, 8 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

OpenAI MiniMax Anthropic Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.10 $1.00 $10 $100 $1,000 Cost of the full run, log scale Claude Fable 5.1, 2026-10-01; 10 runs GPT-5.6 Luna, 2026-10-01; 10 runs MiniMax-M3, 2026-10-01; 10 runs gpt-oss-120b, 2026-10-01; 10 runs 0% 20% 40% 60% 80% 100% $0.10 $1.00 $10 $100 $1,000 Cost of the full run, log scale Claude Fable 5.1, 2026-10-01; 10 runs GPT-5.6 Luna, 2026-10-01; 10 runs

Source: Official React Native Evals leaderboard

ModelEffortHarnessScoreCost of the full runOutput tokens
Claude Fable 5.1 2026-10-01; 10 runs 89% $92
Claude Opus 5.5 2026-10-01; 10 runs 88.6% $29
Claude Opus 5 2026-10-01; 10 runs 88.1% $142
Claude Fable 5 2026-10-01; 10 runs 87% $137
GPT-6 Astra 2026-10-01; 10 runs 86.4% $65
GPT-5.6 Sol 2026-10-01; 10 runs 86.3% $51
GPT-6.1 Sol 2026-10-01; 10 runs 86.3% $12
Claude Opus 4.8 2026-10-01; 10 runs 84.7% $64
GPT-5.5 2026-10-01; 10 runs 84.7% $58
Grok 4.5 2026-10-01; 10 runs 84.6% $43
Grok 4.6 2026-10-01; 10 runs 84.1% $43
GPT-5.6 Terra 2026-10-01; 10 runs 83.5% $17
Claude Sonnet 5 2026-10-01; 10 runs 83.4% $99
GPT-6 Sol 2026-10-01; 10 runs 83.3% $21
GPT-5.6 Luna 2026-10-01; 10 runs 82.9% $4.57
Muse Spark 1.1 2026-10-01; 10 runs 81.1% $19
Gemini 3.1 Pro 2026-10-01; 10 runs 78.9% $49
MiniMax-M3 2026-10-01; 10 runs 78.6% $4.44
GLM-5.2 2026-10-01; 10 runs 78.3% $92
Kimi K2.6 2026-10-01; 10 runs 77.7%
Gemini 3.5 Flash 2026-10-01; 10 runs 75.5%
MiMo-V2.5-Pro 2026-10-01; 10 runs 73.4% $6.25
DeepSeek-V4-Flash 2026-10-01; 10 runs 73.2%
gpt-oss-120b 2026-10-01; 10 runs 71.6% $1.94
gpt-oss-20b 2026-10-01; 10 runs 71% $0.79
Mistral Large 3 2026-10-01; 10 runs 65.5% $13
Nemotron 3 Nano 30B-A3B 2026-10-01; 10 runs 42.7%

6 models have held number one on the official leaderboard since its release in March 2026

Lab claims count from the day the card came out, independent runs from the model’s release.

Google DeepMind OpenAI Anthropic
Own leaderboard Passed 85% in Jun 2026, by Claude Fable 5

Source: Official React Native Evals leaderboard

Claude Fable 5 on the official leaderboard passed 85% in June 2026

That was 3 months after its release. By the same age, 23% of the benchmarks from 2026 had passed 85%. All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Fable 5.1 Claude Fable 5 Gemini 3.1 Pro gpt-oss-120b 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Fable 5.1 Gemini 3.1 Pro

Source: Official React Native Evals leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Coding

All 523 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471

Data audit

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (32 entries).
  • Pass Models scored 27 models we track, plus 5 other entries on its leaderboard.
  • Pass Newest score The newest is from September 2026, for GPT-6.1 Sol.
  • Pass Frontier models 4 of the top 10 have a score, led by GPT-6 Astra (#4).
  • Note Headroom Saturated. The best is 89%, past 85% since June 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on rn-evals.vercel.app: 27 rows for models we track and 5 other entries at the last read.
  • Pass Last read November 18th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check