Benchmark Almanac

Vals Multimodal Index v1.1

Claude Opus 4.8 70.7% leads since May 2026
Not in model cards
Version 1.1

Models complete finance, coding, and education tasks that contribute to a weighted index.

The index uses Finance Agent v2 in finance and Terminal-Bench 2.0 in coding.

Made by
Vals AI
Released
May 2026
Kind
Composite work indices
Level
Expert
Scored
Settled by events
Answers
Agent task
Tasks
Kept private
Measures
Capability
Language
English
In and out
Text, Code, Documents, Images → Text, Code, Actions
Interaction
Agentic
Environment
APIs and tools, Terminal
Items from
Another benchmark
Run by
A host
Combines
An average of other benchmarks
Strongest model tested
GPT-5.5 #21 on the capability index
Last model added
May 2026
Scores refreshed
November 20th, 2026
15
Models scored on its own leaderboard

Claude Opus 4.8 leads with 70.7% among models released by May 2026

Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.

The newest model with a score here came out in May 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 Opus 4.8 70.7%
  2. 2 GPT-5.5 67.8%
  3. 3 Claude Opus 4.7 67.4%
  4. 4 Gemini 3.5 Flash 62.3%
  5. 5 Claude Sonnet 4.6 60.8%
  6. 6 Kimi K2.6 56.8%
  7. 7 Gemini 3.1 Pro 55.8%
  8. 8 GPT-5.4 Mini 53.3%
  9. 9 Gemini 3 Flash 52%
  10. 10 Qwen 3.6 Plus 50.7%
  11. 11 GPT-5.4 Nano 47.5%
  12. 12 Grok 4.3 43.4%
  13. 13 Claude Haiku 4.5 thinking 42.4%
  14. 14 Gemini 3.1 Flash-Lite 40.5%
  15. 15 Grok 4.20 reasoning 38.7%

Source: Official Vals Multimodal Index v1.1 leaderboard

#ModelScore ±SettingDateSource
1 Claude Opus 4.8 70.7% ±1.5 May 2026 Official leaderboard
2 GPT-5.5 67.8% ±2.9 Apr 2026 Official leaderboard
3 Claude Opus 4.7 67.4% ±2.4 Apr 2026 Official leaderboard
4 Gemini 3.5 Flash 62.3% ±2.9 May 2026 Official leaderboard
5 Claude Sonnet 4.6 60.8% ±2.9 Feb 2026 Official leaderboard
6 Kimi K2.6 56.8% ±3.5 Apr 2026 Official leaderboard
7 Gemini 3.1 Pro 55.8% ±2.9 Feb 2026 Official leaderboard
8 GPT-5.4 Mini 53.3% ±3.6 Mar 2026 Official leaderboard
9 Gemini 3 Flash 52% ±2.9 Dec 2025 Official leaderboard
10 Qwen 3.6 Plus 50.7% ±2.8 Mar 2026 Official leaderboard
11 GPT-5.4 Nano 47.5% ±3.4 Mar 2026 Official leaderboard
12 Grok 4.3 43.4% ±2.5 Apr 2026 Official leaderboard
13 Claude Haiku 4.5 42.4% ±2 thinking Oct 2025 Official leaderboard
14 Gemini 3.1 Flash-Lite 40.5% ±1.5 Mar 2026 Official leaderboard
15 Grok 4.20 38.7% ±1.7 reasoning Feb 2026 Official leaderboard

The best score, 70.7%, costs $1.95 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.

Google DeepMind OpenAI Moonshot AI Anthropic 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 Claude Opus 4.8 Gemini 3.5 Flash Kimi K2.6 Gemini 3 Flash GPT-5.4 Nano Gemini 3.1 Flash-Lite 0% 20% 40% 60% 80% 100% $0.10 $1.00 $10 Cost per task, log scale Claude Opus 4.8 Gemini 3.5 Flash

Source: Official Vals Multimodal Index v1.1 leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Opus 4.8 70.7% $1.95
GPT-5.5 67.8% $3.47
Claude Opus 4.7 67.4% $4.33
Gemini 3.5 Flash 62.3% $1.06
Claude Sonnet 4.6 60.8% $1.57
Kimi K2.6 56.8% $0.52
Gemini 3.1 Pro 55.8% $0.97
GPT-5.4 Mini 53.3% $0.56
Gemini 3 Flash 52% $0.29
Qwen 3.6 Plus 50.7% $1.03
GPT-5.4 Nano 47.5% $0.26
Grok 4.3 43.4% $0.52
Claude Haiku 4.5 thinking 42.4% $0.40
Gemini 3.1 Flash-Lite 40.5% $0.12
Grok 4.20 reasoning 38.7% $0.36

2 models have held number one on the official leaderboard since its release in May 2026

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

OpenAI Anthropic
Own leaderboard Not tested since May 2026

Source: Official Vals Multimodal Index v1.1 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 the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Opus 4.8 Claude Sonnet 4.6 Gemini 3 Flash Claude Haiku 4.5 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Opus 4.8 Claude Sonnet 4.6

Source: Official Vals Multimodal Index v1.1 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.

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.

Anthropic OpenAI Google DeepMind
Vals Multimodal Index v1.1 Not tested since May 2026
Vals Multimodal Index v1 Replaced by Index, May 2026
Vals Multimodal Index Replaced by v1.1, Aug 2026
Vals Multimodal Index v1.1 Not tested since May 2026

Source: Official leaderboards

More in Economic work

All 210 in Economic work
  • JobBench In use 130 tasks from 35 occupations workers most want delegated. Agent works through messy reference files, graded by fact-anchored binary rubrics. GLM-5.3 61.4%
  • GDPval In use Real work tasks from 44 occupations with reference files. Experts blindly compare model deliverables with expert work. Metric is win-or-tie rate. Seed2.1 Pro 87.9%
  • GDPval-AA v2.1 In use Agents with shell and web access complete 220 GDPval gold tasks from 44 occupations. Blind pairwise comparisons give Elo ratings. Claude Opus 5.5 1,846
  • AA-Briefcase v1.1 In use Version 1.1 of Artificial Analysis's multi-hour office work evaluation, with Elo ratings fitted by Crowd-BT. Claude Opus 5.5 1,822
  • Agents' Last Exam v1.1 In use Agents complete 122 public long-horizon professional workflow tasks across 55 sub-industries. Deterministic graders check outcomes against hidden references. GPT-6.2 Astra 47.5%
  • CoWorkBench In use Agent completes long-horizon productivity tasks across computer science, finance, law and medicine. Qwen in-house scoring (method unpublished). Claude Fable 5 75.9%

Data audit

All 9 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access The tasks are private.
  • Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (15 entries).
  • Pass Models scored 15 models we track.
  • Note Newest score The newest came 6 months ago, in May 2026, for Claude Opus 4.8. That’s expected for a test replaced by Vals Multimodal Index.
  • Note Frontier models None of the top 10 has a score, which is fine for a retired test. The strongest tested is GPT-5.5 (#21).
  • Pass Headroom The best is 70.7%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on vals.ai: 15 rows for models we track at the last read.
  • Pass Last read November 20th, 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