Best AI models for automation.
Prices checked
For automation, DeepSeek V4.1 Flash 1017 is the best measured match of 42 models with published results, from $0.13 per M tokens at ZenMux, checked .
Top 10
Add your requirements| Rank | Name | Price per M tokens | Links |
|---|---|---|---|
| 1 | DeepSeek V4.1 Flash 1017Best match
| $0.13/M tokens | |
| 2 | Qwen3.8 27BLowest cost
| $0.12/M tokens | |
| 3 | Claude Fable 5.1
| $20/M tokens | |
| 4 | Muse Spark 1.3
| $0.97/M tokens | |
| 5 | Claude Fable 5
| $20/M tokens | |
| 6 | GPT-6 Astra
| $19/M tokens | |
| 7 | Qwen3.8 Max
| $1.95/M tokens | |
| 8 | Kimi K3
| $5.10/M tokens | |
| 9 | GPT-5.6 Sol
| $8/M tokens | |
| 10 | Gemini 3.7 Flash
| $1.50/M tokens |
How hard is the task?
Rank again for a quicker or a harder version of it.
Change the range
Quick tasks
Choose a tool for a well-defined request and populate its arguments.
This list
Balanced
Chain a few tool calls, use their results and handle a routine missing field.
Change the range
Complex tasks
Coordinate several steps with dependencies, recover from tool failures and revise a plan as results arrive.
How we count
An automated workflow asks a model to choose a next step, call a tool and respond to what comes back. This shortlist includes models with confirmed tool support and emphasizes instruction following, agentic coding and reasoning. The surrounding workflow determines which actions are available.
The list weighs the published results that matter for automation and keeps the models close enough to the best one for a balanced task. Price picks the provider after that; it never lifts a weaker model into the list. 325 available models were considered and 42 have comparable evidence.
Each price is the cheapest provider read in the last three days, per M tokens, blended 3 in : 1 out.
This ranking is a weighted capability proxy, not a measured workflow completion rate. Tool support alone does not establish reliable planning or compatibility with your application.
The ranking uses the same Balanced settings as the Ask page. The Capability methodology has the sources, the exact identity rules, the task weights and the difficulty ranges.
- Best match
- The strongest published results for this task at a balanced difficulty.
- Best value
- The cheapest model within five points of the best match.
- Lowest cost
- The cheapest model in the range this page lists.