Benchmark Almanac

Vals Smith: apache/spark

GPT-5.6 Sol 84.5% leads since August 2026
Not in model cards

Agents modify apache/spark to complete software changes from merged pull requests. Hidden tests check the changes and test for regressions.

Vals Smith generates the tasks and publishes model scores for apache/spark in its public repository directory.

Made by
Vals AI
Released
August 2026
Kind
Fixing real repositories
Level
Expert
Scored
Checked automatically
Answers
Agent task
Measures
Capability
Language
English
In and out
Text, Code → Code, Actions
Interaction
Agentic
Environment
Code sandbox
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
In views
Agents
Strongest model tested
Claude Fable 5 #12 on the capability index
Last model added
July 2026
Scores refreshed
November 21st, 2026
14
Entries on Vals AI’s leaderboard, 13 models we track

GPT-5.6 Sol leads with 84.5% among models released by July 2026 GPT-5.6 Sol leads with 84.5% among models released by July 2026

Best score per entry from Vals AI’s leaderboard, read on November 21st, 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 July 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 GPT-5.6 Sol 84.5%
  2. 2 Claude Opus 4.8 82.8%
  3. 3 Claude Fable 5 82.8%
  4. 4 GLM-5.2 81%
  5. 5 Grok 4.5 81%
  6. 6 Kimi K3 81%
  7. 7 Claude Sonnet 5 77.6%
  8. 8 Claude Opus 4.7 75.9%
  9. 9 GPT-5.6 Luna 70.7%
  10. 10 GPT-5.5 65.5%
  11. 11 Gemini 3.5 Flash 65.5%
  12. 12 GPT-5.6 Terra 65.5%
  13. 13 Gemini 3.1 Pro Preview (02/26) 56.9%
  14. 14 Claude Haiku 4.5 nonthinking 56.9%

Source: Vals AI’s leaderboard

#EntryScoreSettingDateSource
1 GPT-5.6 Sol 84.5% Mini-SWE-agent Jul 2026 Official leaderboard
2 Claude Opus 4.8 82.8% Mini-SWE-agent May 2026 Official leaderboard
3 Claude Fable 5 82.8% Mini-SWE-agent Jun 2026 Official leaderboard
4 GLM-5.2 81% Mini-SWE-agent Jun 2026 Official leaderboard
5 Grok 4.5 81% Mini-SWE-agent Jul 2026 Official leaderboard
6 Kimi K3 81% Mini-SWE-agent Jul 2026 Official leaderboard
7 Claude Sonnet 5 77.6% Mini-SWE-agent Jun 2026 Official leaderboard
8 Claude Opus 4.7 75.9% Mini-SWE-agent Apr 2026 Official leaderboard
9 GPT-5.6 Luna 70.7% Mini-SWE-agent Jul 2026 Official leaderboard
10 GPT-5.5 65.5% Mini-SWE-agent Apr 2026 Official leaderboard
11 Gemini 3.5 Flash 65.5% Mini-SWE-agent May 2026 Official leaderboard
12 GPT-5.6 Terra 65.5% Mini-SWE-agent Jul 2026 Official leaderboard
13 Gemini 3.1 Pro Preview (02/26) 56.9% Mini-SWE-agent Official leaderboard
14 Claude Haiku 4.5 56.9% nonthinking, Mini-SWE-agent Oct 2025 Official leaderboard
#ModelScoreSettingDateSource
1 GPT-5.6 Sol 84.5% Mini-SWE-agent Jul 2026 Official leaderboard
2 Claude Opus 4.8 82.8% Mini-SWE-agent May 2026 Official leaderboard
3 Claude Fable 5 82.8% Mini-SWE-agent Jun 2026 Official leaderboard
4 GLM-5.2 81% Mini-SWE-agent Jun 2026 Official leaderboard
5 Grok 4.5 81% Mini-SWE-agent Jul 2026 Official leaderboard
6 Kimi K3 81% Mini-SWE-agent Jul 2026 Official leaderboard
7 Claude Sonnet 5 77.6% Mini-SWE-agent Jun 2026 Official leaderboard
8 Claude Opus 4.7 75.9% Mini-SWE-agent Apr 2026 Official leaderboard
9 GPT-5.6 Luna 70.7% Mini-SWE-agent Jul 2026 Official leaderboard
10 GPT-5.5 65.5% Mini-SWE-agent Apr 2026 Official leaderboard
11 Gemini 3.5 Flash 65.5% Mini-SWE-agent May 2026 Official leaderboard
12 GPT-5.6 Terra 65.5% Mini-SWE-agent Jul 2026 Official leaderboard
13 Claude Haiku 4.5 56.9% nonthinking, Mini-SWE-agent Oct 2025 Official leaderboard

The best score, 84.5%, costs $2.46 a task

Runs on the cost frontier score higher than any run that's cheaper. Grok 4.5 comes within 3.5 points of the best for $0.71 a task, 3 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

xAI OpenAI Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.10 $0.20 $0.50 $1.00 $2.00 $5.00 $10 Cost per task, log scale GPT-5.6 Sol Grok 4.5 GPT-5.6 Luna 0% 20% 40% 60% 80% 100% $0.10 $1.00 $10 Cost per task, log scale GPT-5.6 Sol Grok 4.5

Source: Vals AI’s leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-5.6 Sol Mini-SWE-agent 84.5% $2.46
Claude Opus 4.8 Mini-SWE-agent 82.8% $4.55
Claude Fable 5 Mini-SWE-agent 82.8% $7.44
GLM-5.2 Mini-SWE-agent 81% $0.88
Grok 4.5 Mini-SWE-agent 81% $0.71
Kimi K3 Mini-SWE-agent 81% $9.82
Claude Sonnet 5 Mini-SWE-agent 77.6% $4.96
Claude Opus 4.7 Mini-SWE-agent 75.9% $4.92
GPT-5.6 Luna Mini-SWE-agent 70.7% $0.12
GPT-5.5 Mini-SWE-agent 65.5% $2.02
Gemini 3.5 Flash Mini-SWE-agent 65.5% $1.26
GPT-5.6 Terra Mini-SWE-agent 65.5% $0.48
Claude Haiku 4.5 nonthinking Mini-SWE-agent 56.9% $0.41

The best score, 84.5%, costs $2.46 a task

Runs on the cost frontier score higher than any run that's cheaper. Grok 4.5 comes within 3.5 points of the best for $0.71 a task, 3 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

xAI OpenAI Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.10 $0.20 $0.50 $1.00 $2.00 $5.00 $10 Cost per task, log scale GPT-5.6 Sol Grok 4.5 GPT-5.6 Luna 0% 20% 40% 60% 80% 100% $0.10 $1.00 $10 Cost per task, log scale GPT-5.6 Sol Grok 4.5

Source: Vals AI’s leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-5.6 Sol Mini-SWE-agent 84.5% $2.46
Claude Opus 4.8 Mini-SWE-agent 82.8% $4.55
Claude Fable 5 Mini-SWE-agent 82.8% $7.44
GLM-5.2 Mini-SWE-agent 81% $0.88
Grok 4.5 Mini-SWE-agent 81% $0.71
Kimi K3 Mini-SWE-agent 81% $9.82
Claude Sonnet 5 Mini-SWE-agent 77.6% $4.96
Claude Opus 4.7 Mini-SWE-agent 75.9% $4.92
GPT-5.6 Luna Mini-SWE-agent 70.7% $0.12
GPT-5.5 Mini-SWE-agent 65.5% $2.02
Gemini 3.5 Flash Mini-SWE-agent 65.5% $1.26
GPT-5.6 Terra Mini-SWE-agent 65.5% $0.48
Claude Haiku 4.5 nonthinking Mini-SWE-agent 56.9% $0.41

Number one over time

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

OpenAI
Vals AI leaderboard Not tested since Aug 2026
GPT-5.6 Sol 84.5%

Source: Vals AI’s leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On Vals AI’s leaderboard Best so far
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) GPT-5.6 Sol Claude Opus 4.7 Claude Haiku 4.5 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) GPT-5.6 Sol Claude Opus 4.7

Source: Vals AI’s 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

9 of 10 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark

Catalog record 1 warning 1 pass
  • Warning Task access We don’t know whether the tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (14 entries).
  • Pass Models scored 13 models we track, plus 1 other entry on its leaderboard.
  • Pass Newest score The newest is from July 2026, for Kimi K3.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Claude Fable 5 (#12).
  • Pass Headroom The best is 84.5%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on Vals AI: 13 rows for models we track and 1 other entry at the last read.
  • Pass Last read November 21st, 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