.TXT Serve models?→Release models?↓
Updated October 29, 2026

Tool-call index

Not all models are served equally by all providers. On a typical model, the worst provider breaks tool calls 19× as often as the best. Find the best one for your agents.

← All models
Qwen · 21 providers · 1.3M tool calls a day
#ProviderBroken
tool calls
Tool calls
a day
$ in / out
per M
Tokens/sUptime
1Darkbloomfp4#0.07%3.9k$0.04 / $2.204398.9%
2Venicefp8#0.26%43k$0.45 / $3.20100100.0%
3ModelRun [by Modular]fp4#0.28%116k$0.20 / $2.5013999.6%
4CoreWeavefp8#0.29%94k$0.40 / $3.004699.9%
5DekaLLM#0.40%70k$0.04 / $2.507199.9%
6Cloudflare#0.40%1.4k$0.45 / $3.204299.7%
7Reka AIfp8#0.62%28k$0.04 / $2.2045100.0%
8DeepInfrabf16#0.66%129k$0.15 / $1.884199.6%
9Phala#0.86%47k$0.20 / $2.506799.1%
10Mancerfp8#0.93%1.5k$0.18 / $2.505695.4%
11Chutesfp8#1.00%245k$0.04 / $1.252899.9%
12AkashMLfp8#1.00%9.2k$0.23 / $1.984399.1%
13NEAR AIfp8#1.00%112k$0.04 / $1.352499.8%
14Parasailfp8#1.02%30k$0.24 / $2.208899.9%
15Waferfp8#1.06%62k$0.06 / $2.306399.8%
16Tenstorrentfp4#1.19%15k$0.15 / $1.506294.0%
17Modal#1.38%124k$0.04 / $2.30124100.0%
18Ionstreamfp8#1.49%15k$0.04 / $1.954398.9%
19Alibaba Cloud Int.#2.44%84k$0.42 / $2.554397.5%
20NovitaAI#4.41%14k$0.42 / $3.004199.7%
21Cerebrasfp16#13.98%17k$0.99 / $1.4949499.8%
Recommended OpenRouter settings for Qwen3.8 27B
{
  "provider": {
    "order": [
      "darkbloom/fp4",
      "venice/fp8",
      "modelrun/fp4"
    ],
    "ignore": [
      "novita",
      "cerebras/fp16"
    ]
  }
}

Point your agent at llms.txt

Using Qwen3.8 27B?

Get updates when its successor drops

And which provider to use.

For model makers

Broken tool calls carry your model's name

Releasing a model? Test it before launch.

Worst provider13.98%

of Qwen3.8 27B's tool calls fail on Cerebras (fp16).

Best to worst140×

as often on Cerebras (fp16) as on Darkbloom (fp4), its best provider.

Every day14k

broken tool calls a day across its 21 providers.