Code Migration v1.1
Version 1.1
The official leaderboard hasn’t scored a model released after Claude Fable 5.2 (November 2026). Labs report their own numbers for one newer model: Inkling 2 Mini at 23.2%.
Models port a working program into another language offline and ship a Dockerfile. A hidden behavior-test suite the model never sees scores the port, after an anti-cheat check.
Same tasks as v1, but every model runs on the Pi agent harness and grading uses revised hidden test suites. Vals says scores are not directly comparable to v1. Part of Vals Index v2.2.
- Made by
- Vals AI
- Released
- October 2026
- Kind
- Fixing real repositories
- Level
- Expert
- Scored
- Checked automatically
- Answers
- Code
- Tasks
- Kept private
- Measures
- Capability
- Language
- English
- Interaction
- Agentic
- Environment
- Terminal
- Items from
- Another benchmark
- Refresh
- New versions replace it
- Run by
- A host
- Based on
- Code Migration
- In views
- Agents
- Strongest model tested
- Claude Fable 5.2 #2 on the capability index
- Last in a system card
- November 2026
- Last model added
- November 2026
- Scores refreshed
- November 19th, 2026
Claude Sonnet 5.5 leads with 86.9%
Best score per model from the benchmark's own leaderboard, read on November 19th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.
Claude Sonnet 5.5 leads with 86.9%, claims included
Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model. Striped bars are claims, and dotted bars are reruns by Epoch AI or another board.
- 86.9 Claude Sonnet 5.5
- 83.3 Claude Fable 5.2
- 81.0 Claude Opus 5.5
- 80.6 GPT-6 Astra
- 77.3 GPT-6.1 Sol
- 72.4
Muse Wonder
- 70.8
Gemini 4 Argon
- 67.5
DeepSeek V4.1 Flash 1017
- 62.8
Muse Wonder Full
- 58.9
DeepSeek V4.1 Flash
- 57.2
Qwen3.8 Max (0902)
- 55.1 GLM-5.3
- 53.2 Grok 4.7
- 52.4 Claude Haiku 5.5
- 52.0
DeepSeek V4 Pro 0813
- 51.6 GLM-5.3-Flash
- 48.8 MiMo-V2.6 Pro
- 47.0
Step 5 Preview
- 46.7 GPT-6 Luna
- 45.4
Tencent Hy4 preview
- 39.5 MiMo-V2.6 Flash
- 39.3 Mercury 3
- 38.6
Gemini 3.8 Flash
- 28.6 Kimi K3
- 26.9
Muse Spark Open
- 22.9 Beam 1.0
- 20.9
MiniMax-M3
- 16.4
Nemotron 3.5 Super VL
- 10.4
Inkling-Small
- 8.82
Inkling
- 1 Claude Sonnet 5.5 86.9%
- 2 Claude Fable 5.2 83.3%
- 3 Claude Opus 5.5 81%
- 4 GPT-6 Astra 80.6%
- 5 GPT-6.1 Sol 77.3%
- 6
Muse Wonder 72.4%
- 7
Gemini 4 Argon 70.8%
- 8
DeepSeek V4.1 Flash 1017 67.5%
- 9
Muse Wonder Full 62.8%
- 10
DeepSeek V4.1 Flash 58.9%
- 11
Qwen3.8 Max (0902) 57.2%
- 12 GLM-5.3 55.1%
- 13 Grok 4.7 53.2%
- 14 Claude Haiku 5.5 52.4%
- 15
DeepSeek V4 Pro 0813 52%
- 16 GLM-5.3-Flash 51.6%
- 17 MiMo-V2.6 Pro 48.8%
- 18
Step 5 Preview 47%
- 19 GPT-6 Luna 46.7%
- 20
Tencent Hy4 preview 45.4%
- 21 MiMo-V2.6 Flash 39.5%
- 22 Mercury 3 39.3%
- 23
Gemini 3.8 Flash 38.6%
- 24 Kimi K3 28.6%
- 25
Muse Spark Open 26.9%
- 26 Beam 1.0 22.9%
- 27
MiniMax-M3 20.9%
- 28
Nemotron 3.5 Super VL 16.4%
- 29
Inkling-Small 10.4%
- 30
Inkling 8.82%
Source: Official Code Migration v1.1 leaderboard
- 86.9 Claude Sonnet 5.5
- 83.3 Claude Fable 5.2
- 81.0 Claude Opus 5.5
- 80.6 GPT-6 Astra
- 77.3 GPT-6.1 Sol
- 72.4
Muse Wonder
- 70.8
Gemini 4 Argon
- 67.5
DeepSeek V4.1 Flash 1017
- 62.8
Muse Wonder Full
- 58.9
DeepSeek V4.1 Flash
- 57.2
Qwen3.8 Max (0902)
- 55.1 GLM-5.3
- 53.2 Grok 4.7
- 52.4 Claude Haiku 5.5
- 52.0
DeepSeek V4 Pro 0813
- 51.6 GLM-5.3-Flash
- 48.8 MiMo-V2.6 Pro
- 47.0
Step 5 Preview
- 46.7 GPT-6 Luna
- 45.4
Tencent Hy4 preview
- 39.5 MiMo-V2.6 Flash
- 39.3 Mercury 3
- 38.6
Gemini 3.8 Flash
- 28.6 Kimi K3
- 26.9
Muse Spark Open
- 23.2
Inkling 2 Mini
- 22.9 Beam 1.0
- 20.9
MiniMax-M3
- 16.4
Nemotron 3.5 Super VL
- 10.4
Inkling-Small
- 8.82
Inkling
- 1 Claude Sonnet 5.5 86.9%
- 2 Claude Fable 5.2 83.3%
- 3 Claude Opus 5.5 81%
- 4 GPT-6 Astra 80.6%
- 5 GPT-6.1 Sol 77.3%
- 6
Muse Wonder 72.4%
- 7
Gemini 4 Argon 70.8%
- 8
DeepSeek V4.1 Flash 1017 67.5%
- 9
Muse Wonder Full 62.8%
- 10
DeepSeek V4.1 Flash 58.9%
- 11
Qwen3.8 Max (0902) 57.2%
- 12 GLM-5.3 55.1%
- 13 Grok 4.7 53.2%
- 14 Claude Haiku 5.5 52.4%
- 15
DeepSeek V4 Pro 0813 52%
- 16 GLM-5.3-Flash 51.6%
- 17 MiMo-V2.6 Pro 48.8%
- 18
Step 5 Preview 47%
- 19 GPT-6 Luna 46.7%
- 20
Tencent Hy4 preview 45.4%
- 21 MiMo-V2.6 Flash 39.5%
- 22 Mercury 3 39.3%
- 23
Gemini 3.8 Flash 38.6%
- 24 Kimi K3 28.6%
- 25
Muse Spark Open 26.9%
- 26
Inkling 2 Mini 23.2%
- 27 Beam 1.0 22.9%
- 28
MiniMax-M3 20.9%
- 29
Nemotron 3.5 Super VL 16.4%
- 30
Inkling-Small 10.4%
- 31
Inkling 8.82%
Sources: Official Code Migration v1.1 leaderboard and labs’ model cards
| # | Model | Score ± 95% CI | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Sonnet 5.5 | 86.9% ±3 | Sep 2026 | Official leaderboard | |
| 2 | Claude Fable 5.2 | 83.3% ±6.1 | Nov 2026 | Official leaderboard | |
| 3 | Claude Opus 5.5 | 81% ±6.3 | Sep 2026 | Official leaderboard | |
| 4 | GPT-6 Astra | 80.6% ±4.6 | Sep 2026 | Official leaderboard | |
| 5 | GPT-6.1 Sol | 77.3% ±5.8 | Sep 2026 | Official leaderboard | |
| 6 | | 72.4% ±6.2 | Oct 2026 | Official leaderboard | |
| 7 | | 70.8% ±7.3 | Sep 2026 | Official leaderboard | |
| 8 | | 67.5% ±5.4 | Oct 2026 | Official leaderboard | |
| 9 | | 62.8% ±6.2 | Oct 2026 | Official leaderboard | |
| 10 | | 58.9% ±5.2 | Sep 2026 | Official leaderboard | |
| 11 | | 57.2% ±5.5 | Sep 2026 | Official leaderboard | |
| 12 | GLM-5.3 | 55.1% ±4.1 | Aug 2026 | Official leaderboard | |
| 13 | Grok 4.7 | 53.2% ±5 | Sep 2026 | Official leaderboard | |
| 14 | Claude Haiku 5.5 | 52.4% ±6.8 | Oct 2026 | Official leaderboard | |
| 15 | | 52% ±4.4 | Aug 2026 | Official leaderboard | |
| 16 | GLM-5.3-Flash | 51.6% ±5.9 | Aug 2026 | Official leaderboard | |
| 17 | MiMo-V2.6 Pro | 48.8% ±7.6 | Sep 2026 | Official leaderboard | |
| 18 | | 47% ±6.7 | Sep 2026 | Official leaderboard | |
| 19 | GPT-6 Luna | 46.7% ±7.1 | Sep 2026 | Official leaderboard | |
| 20 | | 45.4% ±7 | Aug 2026 | Official leaderboard | |
| 21 | MiMo-V2.6 Flash | 39.5% ±6.9 | Sep 2026 | Official leaderboard | |
| 22 | Mercury 3 | 39.3% ±6.2 | Oct 2026 | Official leaderboard | |
| 23 | | 38.6% ±7.2 | Sep 2026 | Official leaderboard | |
| 24 | Kimi K3 | 28.6% ±7.4 | Jul 2026 | Official leaderboard | |
| 25 | | 26.9% ±6.3 | Oct 2026 | Official leaderboard | |
| 26 | Beam 1.0 | 22.9% ±6.4 | Oct 2026 | Official leaderboard | |
| 27 | | 20.9% ±6.7 | Jun 2026 | Official leaderboard | |
| 28 | | 16.4% ±5.8 | Oct 2026 | Official leaderboard | |
| 29 | | 10.4% ±4.6 | Jul 2026 | Official leaderboard | |
| 30 | | 8.82% ±5.3 | Jul 2026 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Sonnet 5.5 | 86.9% | Sep 2026 | Official leaderboard | |
| 2 | Claude Fable 5.2 | 83.3% | Nov 2026 | Official leaderboard | |
| 3 | Claude Opus 5.5 | 81% | Sep 2026 | Official leaderboard | |
| 4 | GPT-6 Astra | 80.6% | Sep 2026 | Official leaderboard | |
| 5 | GPT-6.1 Sol | 77.3% | Sep 2026 | Official leaderboard | |
| 6 | 72.4% | Oct 2026 | Official leaderboard | ||
| 7 | 70.8% | Sep 2026 | Official leaderboard | ||
| 8 | 67.5% | Oct 2026 | Official leaderboard | ||
| 9 | 62.8% | Oct 2026 | Official leaderboard | ||
| 10 | 58.9% | Sep 2026 | Official leaderboard | ||
| 11 | 57.2% | Sep 2026 | Official leaderboard | ||
| 12 | GLM-5.3 | 55.1% | Aug 2026 | Official leaderboard | |
| 13 | Grok 4.7 | 53.2% | Sep 2026 | Official leaderboard | |
| 14 | Claude Haiku 5.5 | 52.4% | Oct 2026 | Official leaderboard | |
| 15 | 52% | Aug 2026 | Official leaderboard | ||
| 16 | GLM-5.3-Flash | 51.6% | Aug 2026 | Official leaderboard | |
| 17 | MiMo-V2.6 Pro | 48.8% | Sep 2026 | Official leaderboard | |
| 18 | 47% | Sep 2026 | Official leaderboard | ||
| 19 | GPT-6 Luna | 46.7% | Sep 2026 | Official leaderboard | |
| 20 | 45.4% | Aug 2026 | Official leaderboard | ||
| 21 | MiMo-V2.6 Flash | 39.5% | Sep 2026 | Official leaderboard | |
| 22 | Mercury 3 | 39.3% | Oct 2026 | Official leaderboard | |
| 23 | 38.6% | Sep 2026 | Official leaderboard | ||
| 24 | Kimi K3 | 28.6% | Jul 2026 | Official leaderboard | |
| 25 | 26.9% | Oct 2026 | Official leaderboard | ||
| 26 | 23.2% | Nov 2026 | Lab model card | ||
| 27 | Beam 1.0 | 22.9% | Oct 2026 | Official leaderboard | |
| 28 | 20.9% | Jun 2026 | Official leaderboard | ||
| 29 | 16.4% | Oct 2026 | Official leaderboard | ||
| 30 | 10.4% | Jul 2026 | Official leaderboard | ||
| 31 | 8.82% | Jul 2026 | Official leaderboard |
Labs report scores for 1 model in their cards. Switch the scores above to rank them together.
The best score, 86.9%, costs $44 a task
Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.
Source: Official Code Migration v1.1 leaderboard
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| Claude Sonnet 5.5 | 86.9% | $44 | 1,285,731 | ||
| Claude Fable 5.2 | 83.3% | $69 | 825,042 | ||
| Claude Opus 5.5 | 81% | $72 | 1,011,652 | ||
| GPT-6 Astra | 80.6% | $62 | 217,353 | ||
| GPT-6.1 Sol | 77.3% | $9.51 | 245,105 | ||
| 72.4% | $48 | 428,258 | |||
| 70.8% | $45 | 466,776 | |||
| 67.5% | $1.76 | 439,343 | |||
| 62.8% | $24 | 276,462 | |||
| 58.9% | $1.09 | 350,058 | |||
| 57.2% | $27 | 382,294 | |||
| GLM-5.3 | 55.1% | $35 | 363,586 | ||
| Grok 4.7 | 53.2% | $38 | 189,642 | ||
| Claude Haiku 5.5 | 52.4% | $6.20 | 509,865 | ||
| 52% | $12 | 494,464 | |||
| GLM-5.3-Flash | 51.6% | $8.96 | 496,106 | ||
| MiMo-V2.6 Pro | 48.8% | $0.99 | 355,485 | ||
| 47% | $8.61 | 375,972 | |||
| GPT-6 Luna | 46.7% | $0.83 | 223,402 | ||
| 45.4% | $3.58 | 241,442 | |||
| MiMo-V2.6 Flash | 39.5% | $0.53 | 496,641 | ||
| Mercury 3 | 39.3% | $2.11 | 298,888 | ||
| 38.6% | $16 | 216,779 | |||
| Kimi K3 | 28.6% | $36 | 335,003 | ||
| 26.9% | $3.77 | 90,650 | |||
| Beam 1.0 | 22.9% | 45,577 | |||
| 20.9% | $8.39 | 147,878 | |||
| 16.4% | $1.60 | 193,434 | |||
| 10.4% | $1.36 | 99,516 | |||
| 8.82% | $6.13 | 66,436 |
Number one over time Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official Code Migration v1.1 leaderboard
Claude Sonnet 5.5 on the official leaderboard passed 85% in September 2026
Models released before it came out already scored that high. All the scores we have, one dot each, so a model tested at several settings appears more than once.
Sources: Official Code Migration v1.1 leaderboard and labs’ model cards
Labs report it only in system cards
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark. Some system cards do, below.
One system card reports it
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Nov 20, 2026 | | 23.2 | Code Migration v1.1 (pi, Vals) | Capabilities |
Versions
We switch the top row to each new version when it comes out. Open it for one row per version. An old version stays in play while people still test it, so whoever first takes it past 85% gets credit for closing it out.
Source: Official leaderboards
More 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 The tasks are private.
- Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
- Pass Independent runs Scores from the official leaderboard (30 entries) and system cards (1 score).
- Pass Models scored 31 models we track.
- Pass Newest score The newest is from November 2026, for Inkling 2 Mini.
- Pass Frontier models 6 of the top 10 have a score, led by Claude Fable 5.2 (#2).
- Note Headroom Saturated. The best is 86.9%, past 85% since September 2026.
Leaderboard import 3 pass
- Pass Importer We read the board on vals.ai: 30 rows for models we track at the last read.
- Pass Last read November 19th, 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.