2026-09-10MODELDeepSeek-V4.1-Flash released with extreme agent memory efficiency and evaluated across VALS • 2026-09-03MODELGPT-6 Astra released by OpenAI - 71.7% on EMB, 63.3% on Tax Agent Bench • 2026-09-02MODELGoogle Gemini 3.8 Flash takes #1 rank on Finance Agent (v2) with 61.4% and 72.2% on EMB • 2026-09-02MODELMeta releases Muse Spark 1.3 & 1.3 Max - 59.0%-60.0% on Finance Agent (v2) • 2026-09-01MODELClaude Fable 5.1 leads Tax Agent Bench (77.6%) and EMB (76.7%) with 75% cheaper cache reads • 2026-09-11BENCHTax Agent Bench and FAB v2 benchmarks updated with latest frontier model rankings • 2026-09-15UPDATEPlatform updated: 43 models tracked across 41 independent benchmarks • 2026-09-10MODELDeepSeek-V4.1-Flash released with extreme agent memory efficiency and evaluated across VALS • 2026-09-03MODELGPT-6 Astra released by OpenAI - 71.7% on EMB, 63.3% on Tax Agent Bench • 2026-09-02MODELGoogle Gemini 3.8 Flash takes #1 rank on Finance Agent (v2) with 61.4% and 72.2% on EMB • 2026-09-02MODELMeta releases Muse Spark 1.3 & 1.3 Max - 59.0%-60.0% on Finance Agent (v2) • 2026-09-01MODELClaude Fable 5.1 leads Tax Agent Bench (77.6%) and EMB (76.7%) with 75% cheaper cache reads • 2026-09-11BENCHTax Agent Bench and FAB v2 benchmarks updated with latest frontier model rankings • 2026-09-15UPDATEPlatform updated: 43 models tracked across 41 independent benchmarks •
Home/Benchmarks/SkillsBench
Tier 3agentTasks

SkillsBench

Measures procedural tool orchestration and software skill invocation.

Variable tasks
17 models evaluated
Updated 2026
VALS (2026)
View Source →Submit your scoreMethodology
66.0%
Top Score
Top model
Grok 4.5
Model Rankings
financebenchmark.ai · SkillsBench
Anthropic
Anthropic
OpenAI
DeepSeek
xAI
Zhipu AI
OpenAI
OpenAI
xAI
Thinking Machines
Thinking Machines
Moonshot AI
Ant Group
Meta
Meta
Alibaba
Alibaba
45.1%
🥇AnthropicAnthropicClaude Opus 560.4%
🥈AnthropicAnthropicClaude Sonnet 546.5%
🥉OpenAIOpenAIGPT-5.6 Sol54.1%
4DeepSeekDeepSeekDeepSeek V4 FlashOPEN50.7%
5xAIxAIGrok 4.655.8%
6Zhipu AIZhipu AIGLM 5.2OPEN45.1%
7OpenAIOpenAIGPT-5.6 Luna60.5%
8OpenAIOpenAIGPT-5.6 Terra60.6%
9xAIxAIGrok 4.566.0%
10Thinking MachinesThinking MachinesInklingOPEN26.1%
11Thinking MachinesThinking MachinesInkling SmallOPEN33.6%
12Moonshot AIMoonshot AIKimi K2.7 CodeOPEN50.0%
13Ant GroupAnt GroupLing 3.0 FlashOPEN27.9%
14MetaMetaMuse Spark 1.159.2%
15MetaMetaMuse Spark 1.253.0%
16AlibabaAlibabaQwen 3.7 Plus54.3%
17AlibabaAlibabaQwen 3.8 MaxOPEN42.0%
Full Results Table
#ModelOverallSourceDate
🥇
xAIGrok 4.5
xAI
66.0VALS (2026)2026-07
🥈
OpenAIGPT-5.6 Terra
OpenAI
60.6VALS (2026)2026-07
🥉
OpenAIGPT-5.6 Luna
OpenAI
60.5VALS (2026)2026-07
4
AnthropicClaude Opus 5
Anthropic
60.4VALS (2026)2026-07
5
MetaMuse Spark 1.1
Meta
59.2VALS (2026)2026-07
6
xAIGrok 4.6
xAI
55.8VALS (2026)2026-08
7
AlibabaQwen 3.7 Plus
Alibaba
54.3VALS (2026)2026-06
8
OpenAIGPT-5.6 Sol
OpenAI
54.1VALS (2026)2026-07
9
MetaMuse Spark 1.2
Meta
53.0VALS (2026)2026-08
10
DeepSeekDeepSeek V4 FlashOPEN
DeepSeek
50.7VALS (2026)2026-07
11
Moonshot AIKimi K2.7 CodeOPEN
Moonshot AI
50.0VALS (2026)2026-06
12
AnthropicClaude Sonnet 5
Anthropic
46.5VALS (2026)2026-06
13
Zhipu AIGLM 5.2OPEN
Zhipu AI
45.1VALS (2026)2026-06
14
AlibabaQwen 3.8 MaxOPEN
Alibaba
42.0VALS (2026)2026-08
15
Thinking MachinesInkling SmallOPEN
Thinking Machines
33.6VALS (2026)2026-07
16
Ant GroupLing 3.0 FlashOPEN
Ant Group
27.9VALS (2026)2026-07
17
Thinking MachinesInklingOPEN
Thinking Machines
26.1VALS (2026)2026-07
All evaluated models
Anthropic
Claude Opus 5
60.4%
Anthropic
Claude Sonnet 5
46.5%
OpenAI
GPT-5.6 Sol
54.1%
DeepSeek
DeepSeek V4 Flash
50.7%
xAI
Grok 4.6
55.8%
Zhipu AI
GLM 5.2
45.1%
OpenAI
GPT-5.6 Luna
60.5%
OpenAI
GPT-5.6 Terra
60.6%
xAI
Grok 4.5
66.0%
Thinking Machines
Inkling
26.1%
Thinking Machines
Inkling Small
33.6%
Moonshot AI
Kimi K2.7 Code
50.0%
Ant Group
Ling 3.0 Flash
27.9%
Meta
Muse Spark 1.1
59.2%
Meta
Muse Spark 1.2
53.0%
Alibaba
Qwen 3.7 Plus
54.3%
Alibaba
Qwen 3.8 Max
42.0%