Best AI model to automate tasks with agents
Using tools, querying systems and finishing multi-step tasks without getting lost. Here the model acts, not just answers.
Suggested models
1 Ling 3.1 Flash Ant Group
Why it fits: Completes 79.3% of τ-bench Verified (airline) tasks, which checks whether the model uses systems and follows the rules to the end.
- τ-bench Verified (airline): 79.3%, per OpenRouter Benchmarks, latest version we checked?what it is
- Price: US$ 0.045 input and US$ 0.14 output per 1M tokens (source: OpenRouter)?
- Estimated cost (see the How we calculate block): about US$ 0.0029 (estimate)
- Weights: Undisclosed?
- Alternative: MiMo V2.6 Pro, with open weights, 78.9% on τ-bench Verified (airline)
2 Claude Fable 5.1 Anthropic
Why it fits: Completes 79.2% of τ-bench Verified (airline) tasks, which checks whether the model uses systems and follows the rules to the end.
- τ-bench Verified (airline): 79.2%, per OpenRouter Benchmarks, latest version we checked?
- Price: US$ 10.00 input and US$ 50.00 output per 1M tokens (source: OpenRouter)?
- Estimated cost (see the How we calculate block): about US$ 0.75 (estimate)
- Weights: Closed?
- Alternative: MiMo V2.6 Pro, with open weights, 78.9% on τ-bench Verified (airline)
The latest version, Claude Fable 5.2, does not have a published score yet.
3 MiMo V2.6 Pro Xiaomi
Why it fits: Completes 78.9% of τ-bench Verified (airline) tasks, which checks whether the model uses systems and follows the rules to the end.
- τ-bench Verified (airline): 78.9%, per OpenRouter Benchmarks, latest version we checked?
- Price: US$ 0.43 input and US$ 0.87 output per 1M tokens (source: OpenRouter)?
- Estimated cost (see the How we calculate block): about US$ 0.026 (estimate)
- Weights: Open?
- Alternative: Ling 3.1 Flash, cheaper: US$ 0.045 input per 1M tokens, 79.3% on τ-bench Verified (airline)
4 GLM-5.3 Zhipu AI (Z.ai)
Why it fits: Completes 78.8% of τ-bench Verified (airline) tasks, which checks whether the model uses systems and follows the rules to the end.
- τ-bench Verified (airline): 78.8%, per OpenRouter Benchmarks, latest version we checked?
- Price: US$ 0.07 input and US$ 1.40 output per 1M tokens (source: OpenRouter)?
- Estimated cost (see the How we calculate block): about US$ 0.01 (estimate)
- Weights: Open?
- Alternative: DeepSeek V4.1 Flash, cheaper: US$ 0.022 input per 1M tokens, 77.8% on τ-bench Verified (airline)
Order: highest τ-bench Verified (airline) score among agents and automation models, one version per model line.
How we calculate
Change the numbers and the cost of every model on the page is recalculated right away. Without JavaScript, the default values below apply.
Input = steps × context tokens resent per step. Output = steps × output tokens per step.
With these numbers: about 50,000 input tokens and 5,000 output tokens (estimate).
Conversion used: 1 token is about 4 characters or 0.75 words, OpenAI's rough rule for English text. Other languages can differ. Conversion source: OpenAI, Understanding and counting tokens
Fields marked as assumption are our own guesses, without a source. Adjust them to your case.
Cost of each model = input tokens × input price + output tokens × output price, with prices per 1M tokens. Models that reason before answering can use many more output tokens. It is a reference calculation and does not change the order of the list.
Other paid models we checked
Reference paid models from the main labs. They are left out of the list above because of a lower score, no score on this metric, or not passing the case filter. Prices come from the company's official page when it publishes them.
| Model | Score on this metric | Price and source | Estimated cost |
|---|---|---|---|
| GPT-6 Astra OpenAI | 69.7% | US$ 10.00 input and US$ 50.00 output per 1M tokens (source: OpenAI, official page) | about US$ 0.75 (estimate) |
| GPT-6.1 Sol OpenAI | 70.7% | US$ 2.00 input and US$ 10.00 output per 1M tokens (source: OpenAI, official page) | about US$ 0.15 (estimate) |
| GPT-6 Luna OpenAI | 68.4% | US$ 0.10 input and US$ 0.50 output per 1M tokens (source: OpenAI, official page) | about US$ 0.0075 (estimate) |
| Claude Fable 5.2 Anthropic | no score on this metric | US$ 8.00 input and US$ 40.00 output per 1M tokens (source: Anthropic, official page) | about US$ 0.60 (estimate) |
| Claude Opus 5.5 Anthropic | 75.3% | US$ 4.00 input and US$ 20.00 output per 1M tokens (source: Anthropic, official page) | about US$ 0.30 (estimate) |
| Claude Sonnet 5.5 Anthropic | 78.8% | US$ 2.00 input and US$ 10.00 output per 1M tokens (source: Anthropic, official page) | about US$ 0.15 (estimate) |
| Claude Haiku 5.5 Anthropic | 74% | US$ 0.10 input and US$ 0.50 output per 1M tokens (source: Anthropic, official page; up to 100,000 prompt tokens) | about US$ 0.0075 (estimate) |
| Gemini 4 Argon Google | 78.7% | US$ 2.00 input and US$ 10.00 output per 1M tokens promotional price (source: Google, official page; promotional price currently shown on the page; Google has said this price will go up) | about US$ 0.15 (estimate) |
| Gemini 3.8 Flash Google | 76.7% | US$ 0.75 input and US$ 3.75 output per 1M tokens promotional price (source: Google, official page; promotional price currently shown on the page; Google has said this price will go up) | about US$ 0.056 (estimate) |
| Grok 4.7 xAI | 75.3% | US$ 2.00 input and US$ 6.00 output per 1M tokens (source: xAI, official page) | about US$ 0.13 (estimate) |
| Mistral Large 4 Mistral AI | no score on this metric | US$ 0.50 input and US$ 1.50 output per 1M tokens (source: Mistral AI, official page; price the official page gives for Mistral Large, without naming the version) | about US$ 0.033 (estimate) |
| Mistral Medium 3.5 Mistral AI | no score on this metric | US$ 1.50 input and US$ 7.50 output per 1M tokens (source: OpenRouter; the price could not be read from the official page, because the table is built automatically; OpenRouter is kept as the reference) | about US$ 0.11 (estimate) |
| Command A Cohere | no score on this metric | US$ 2.50 input and US$ 10.00 output per 1M tokens (source: OpenRouter; Cohere's official page does not publish this model's price; OpenRouter is kept as reference) | about US$ 0.17 (estimate) |
| Command A+ Cohere | no score on this metric | US$ 0.30 input and US$ 1.50 output per 1M tokens (source: OpenRouter; Cohere's official page does not publish this model's price; OpenRouter is kept as reference) | about US$ 0.022 (estimate) |
| Nova Premier 1.0 Amazon | no score on this metric | US$ 2.50 input and US$ 12.50 output per 1M tokens (source: Amazon, official page; us-east-1 region, official AWS price list) | about US$ 0.19 (estimate) |
| Nova 2 Lite Amazon | no score on this metric | US$ 0.33 input and US$ 2.75 output per 1M tokens (source: Amazon, official page; us-east-1 region, official AWS price list) | about US$ 0.03 (estimate) |
FAQ
What is the best AI model to automate tasks with agents?
By this page's rule, the list starts with Ling 3.1 Flash, Claude Fable 5.1, MiMo V2.6 Pro, GLM-5.3. Order: highest τ-bench Verified (airline) score among agents and automation models, one version per model line.
How much does it cost to automate tasks with agents with AI?
With the default values in the How we calculate block, the estimate ranges from US$ 0.0029 to US$ 0.75 among the listed models with a published price. It is only an estimate and you can adjust the numbers on the page.
Is there an open weights option to automate tasks with agents?
Yes: MiMo V2.6 Pro, GLM-5.3.
Does this page use AI to recommend?
No. The recommendation comes from fixed rules over public data from Epoch AI, OpenRouter, Hugging Face and each company's official pricing page, in the latest version we checked. The typed text is only compared with keywords.
Terms used on this page
Use cases with their own page
Official pricing pages checked
Sources
- Epoch AI, CC BY 4.0, latest version we checked.
- OpenRouter Benchmarks, CC BY 4.0, latest version we checked.
- OpenRouter, dados públicos da API; exports de benchmarks e rankings em CC BY 4.0, latest version we checked.
- Hugging Face Hub, metadados públicos do Hub; licença de cada modelo no repositório, latest version we checked.
Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch. Family and version were grouped by rule from the name.