Tool Call Total Time (RightRank Probe)

Speed

Time to the full tool call of a request that must call one tool: median of 5 streamed runs through OpenRouter, pinned to the model's first-party provider, from the us-east-1 region.

Lower is better

Rank	Model	Provider	Time
1	Google: Gemini 2.5 Flash LiteRetiring	Google	313 ms p90 353 ms · 5 runs
2	Inception: Mercury 2	Inception	337 ms p90 364 ms · 5 runs
3	Mistral: Ministral 3 3B 2512	Mistral AI	362 ms p90 610 ms · 5 runs
4	Mistral: Mistral Medium 3	Mistral AI	372 ms p90 819 ms · 5 runs
5	Google: Gemini 3.5 Flash Lite	Google	402 ms p90 451 ms · 5 runs
6	Mistral: Mistral Medium 3.1	Mistral AI	404 ms p90 652 ms · 5 runs
7	Mistral: Ministral 3 8B 2512	Mistral AI	436 ms p90 470 ms · 5 runs
8	SpaceXAI: Grok 4.20	SpaceXAI	436 ms p90 662 ms · 5 runs
9	Mistral: Ministral 3 14B 2512	Mistral AI	437 ms p90 819 ms · 5 runs
10	Google: Gemini 2.5 FlashRetiring	Google	438 ms p90 541 ms · 5 runs
11	Mistral: Codestral 2508	Mistral AI	440 ms p90 515 ms · 5 runs
12	Google: Gemini 3.1 Flash Lite Preview	Google	440 ms p90 590 ms · 5 runs
13	OpenAI: GPT-4o-mini	OpenAI	449 ms p90 545 ms · 5 runs
14	Google: Gemini 3.1 Flash Lite	Google	457 ms p90 546 ms · 5 runs
15	Cohere: Command A+	Cohere	463 ms p90 795 ms · 5 runs
16	Qwen: Qwen3 30B A3B Instruct 2507	Alibaba	501 ms p90 537 ms · 5 runs
17	Mistral: Voxtral Small 24B 2507	Mistral AI	518 ms p90 830 ms · 5 runs
18	Amazon: Nova 2 Lite	Amazon	525 ms p90 737 ms · 5 runs
19	Mistral: Devstral 2 2512	Mistral AI	536 ms p90 555 ms · 5 runs
20	Qwen: Qwen3 Next 80B A3B InstructHistorical	Alibaba	539 ms p90 563 ms · 5 runs
21	Amazon: Nova Micro 1.0	Amazon	570 ms p90 766 ms · 5 runs
22	OpenAI: GPT-4o-mini (2024-07-18)	OpenAI	574 ms p90 655 ms · 5 runs
23	Mistral: Mistral Large 3 2512	Mistral AI	580 ms p90 731 ms · 5 runs
24	Qwen: Qwen3 VL 8B InstructHistorical	Alibaba	607 ms p90 768 ms · 5 runs
25	Qwen: Qwen3 VL 30B A3B Instruct	Alibaba	644 ms p90 667 ms · 5 runs
26	OpenAI: GPT-3.5 Turbo	OpenAI	653 ms p90 1.0 s · 5 runs
27	OpenAI: GPT-5.4 Mini	OpenAI	665 ms p90 801 ms · 5 runs
28	Mistral: Mistral Medium 3.5	Mistral AI	671 ms p90 734 ms · 5 runs
29	Amazon: Nova Lite 1.0	Amazon	674 ms p90 760 ms · 5 runs
30	Cohere: North Mini Code (free)	Cohere	689 ms p90 1.0 s · 5 runs
31	Google: Gemini 3 Flash Preview	Google	698 ms p90 793 ms · 5 runs
32	Inception: Mercury 2.5	Inception	701 ms p90 925 ms · 5 runs
33	OpenAI: GPT-6 Luna	OpenAI	708 ms p90 1.0 s · 5 runs
34	Qwen: Qwen3 235B A22B Instruct 2507	Alibaba	711 ms p90 814 ms · 5 runs
35	OpenAI: GPT-4.1 Mini	OpenAI	713 ms p90 725 ms · 5 runs
36	OpenAI: GPT-5.4 Nano	OpenAI	720 ms p90 969 ms · 5 runs
37	Google: Gemini 3.8 Flash	Google	725 ms p90 1.0 s · 5 runs
38	Qwen: Qwen Plus 0728Historical	Alibaba	733 ms p90 842 ms · 5 runs
39	OpenAI: GPT-4o (2024-05-13)	OpenAI	744 ms p90 866 ms · 5 runs
40	OpenAI: GPT-3.5 Turbo 16k	OpenAI	753 ms p90 881 ms · 5 runs
41	OpenAI: GPT-5.6 Luna	OpenAI	756 ms p90 815 ms · 5 runs
42	Qwen: Qwen3 VL 32B InstructHistorical	Alibaba	808 ms p90 1.2 s · 5 runs
43	Qwen: Qwen-Plus	Alibaba	819 ms p90 872 ms · 5 runs
44	OpenAI: GPT-4.1 Nano	OpenAI	828 ms p90 925 ms · 5 runs
45	Amazon: Nova Pro 1.0	Amazon	882 ms p90 1.0 s · 5 runs
46	Google: Gemini 3.6 Flash	Google	915 ms p90 1.0 s · 5 runs
47	Qwen: Qwen3 VL 235B A22B Instruct	Alibaba	933 ms p90 1.1 s · 5 runs
48	Qwen: Qwen3 Coder PlusHistorical	Alibaba	946 ms p90 1.0 s · 5 runs
49	Qwen: Qwen3.5-122B-A10B	Alibaba	957 ms p90 1.1 s · 5 runs
50	Google: Gemini 3.7 Flash	Google	972 ms p90 1.6 s · 5 runs

Access this data

JSON GET /api/v1

[
  {
    "rank": 1,
    "model": "Google: Gemini 2.5 Flash Lite",
    "provider": "Google",
    "score": "312.5"
  },
  {
    "rank": 2,
    "model": "Inception: Mercury 2",
    "provider": "Inception",
    "score": "337.1"
  },
  {
    "rank": 3,
    "model": "Mistral: Ministral 3 3B 2512",
    "provider": "Mistral AI",
    "score": "361.6"
  }
]

Curl

$ curl -s https://rightrank.ai/api/v1/rankings/tool-call-total-time-rightrank-probe.json | jq '.[:3]'

CLI brew install rightrank

$ rightrank top tool-call-total-time-rightrank-probe --limit 5
$ rightrank compare claude-opus-4.6 gpt-5.4-pro --category tool-call-total-time-rightrank-probe
$ rightrank value tool-call-total-time-rightrank-probe  # best performance per dollar

Claude Code Skill Skill

# In Claude Code, use the RightRank skill:
/rightrank best model for tool-call-total-time-rightrank-probe
/rightrank compare top 5 by value
/rightrank cheapest model scoring above 80 on tool-call-total-time-rightrank-probe