Pareto 26.10 Preview overview
Pareto 26.10 Preview by Unbiased
Pareto 26.10 Preview is a multimodal composite model for research, coding and agentic workflows, with a 1,048,576-token context window.
Pareto 26.10 Preview specs
- Model ID
unbiased/pareto-26.10-preview- Developer
- Unbiased
- Context window
- 1,048,576 tokens
- Max output
- 131,072 tokens
- Input
- Image, Text
- Output
- Text
- API
- OpenAI-compatible, base URL
https://api.aimlapi.com/v1
What Is Pareto 26.10 Preview?
Pareto is a composite model from Unbiased: instead of one set of weights, it runs a mix of frontier and open-source models against the same request and returns the best result. The model is chosen per request and does not switch mid-conversation, which is what lets prompt caching keep working across turns.
26.10 is a preview of the next Pareto release. It changes two things that matter: the context window grows from 262,144 to 1,048,576 tokens, and the price falls to roughly a third of the September release.
Technical Reference
- Model ID: unbiased/pareto-26.10-preview
- Context window: 1,048,576 tokens
- Modalities: Image, Text → Text
- Capabilities: streaming, tool calling, vision
- Endpoint: POST /v1/chat/completions
What Unbiased Reports
Unbiased publishes four figures for this release, from runs dated 1 October 2026:
- DeepSWE v1.1 (agentic coding): 69.9%
- Terminal-Bench 4.0 (agentic terminal use): 50.8%
- Humanity's Last Exam, text-only (expert reasoning): 49.9%
- GPQA-Diamond (science reasoning): 92.4%
These are the vendor's own numbers and Unbiased marks them preliminary, “results are from October 1, 2026 runs and may change before final publication.” They have not been independently reproduced.
What Changed From Pareto
Four times the context
1,048,576 tokens against 262,144 on the September release, enough for a whole repository, a large document set or a long agent transcript in one prompt.
Preview status
Unbiased states the preview may change without notice and points to the stable Pareto release for behaviour you can pin.
Where It Makes the Most Sense
Repository-scale coding
Read whole codebases in one prompt and work across files without chunking.
Agentic workflows
Tool calling and streaming suit agent loops that plan, call tools and iterate; the two agentic benchmarks Unbiased reports are the ones it leads with.
Research over long sources
Papers, contracts and report sets fit inside the context window intact.
Multimodal analysis
Charts, screenshots and scanned pages go into the prompt alongside the text around them.
Use cases
- Repository-scale code generation, debugging and review inside a single prompt
- Agentic workflows with tool calling and streaming
- Research and summarisation over long documents without chunking
- Multimodal analysis of screenshots, diagrams and scanned pages
- Evaluating the next Pareto release before it reaches stable
Capabilities
- Streaming
- Function calling
- Vision
Pareto 26.10 Preview API pricing
| Usage | Price | Unit |
|---|---|---|
| Input | $0.8464 | 1M tokens |
| Output | $3.39 | 1M tokens |
| Cached input | $0.0317 | 1M tokens |
Prices from the AI/ML API price list, checked on Nov 4, 2026.
Pareto 26.10 Preview vs other models: token prices
AI/ML API input and output rates for 5 models, per million tokens.
| Model | Input | Output |
|---|---|---|
 Pareto 26.10 Preview | $0.8464 | $3.39 |
| Pareto 26.10 | $0.8464 | $3.39 |
| Nova Pro 1.0 | $0.8464 | $3.39 |
 Qwen 3.8 27B | $0.4761 | $3.39 |
| Pareto | $2.65 | $7.93 |
How to use the Pareto 26.10 Preview API
View docs
Build with your agent
Give your agent the model details, code and documentation.
1Get your API key
Create API keyexport AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"2Make your first request
curl 'https://api.aimlapi.com/v1/chat/completions' \
-H "Authorization: Bearer $AIMLAPI_KEY" \
-H 'Content-Type: application/json' \
-d '{
"model": "unbiased/pareto-26.10-preview",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}'import os
import requests
headers = {
"Authorization": f"Bearer {os.environ['AIMLAPI_KEY']}",
"Content-Type": "application/json",
}
payload = {
"model": "unbiased/pareto-26.10-preview",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}
response = requests.post("https://api.aimlapi.com/v1/chat/completions", headers=headers, json=payload, timeout=120)
response.raise_for_status()
print(response.json())const response = await fetch("https://api.aimlapi.com/v1/chat/completions", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.AIMLAPI_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"model": "unbiased/pareto-26.10-preview",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}),
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
console.log(await response.json());Models similar to Pareto 26.10 Preview
Pareto 26.10 Preview FAQ
What is Pareto 26.10 Preview?
It is a preview release of Unbiased's composite model. Rather than one set of weights, Pareto runs a mix of frontier and open-source models against the same request and returns the best result. The 26.10 preview raises the context window to 1,048,576 tokens, four times the 262,144 of the September release.
Should I use the preview in production?
Unbiased marks this release as a preview that may change without notice, and points to the stable Pareto release for behaviour you can pin. Use the preview to evaluate the larger context window and the lower price; keep production on the stable model until Unbiased promotes this one.
What does Pareto 26.10 Preview cost?
Pareto 26.10 Preview costs $0.8464 input and $3.39 output per 1M tokens on AI/ML API, billed as you go with no subscription.
How does it score on benchmarks?
Unbiased reports 69.9% on DeepSWE v1.1 (agentic coding), 50.8% on Terminal-Bench 4.0 (agentic terminal use), 49.9% on Humanity's Last Exam text-only, and 92.4% on GPQA-Diamond. The runs are dated 1 October 2026 and Unbiased marks them preliminary, they may change before final publication, and they have not been independently reproduced.
What can it take as input?
Text and images in, text out. Screenshots, diagrams and scanned pages can go straight into the prompt alongside the surrounding text. It supports streaming, tool calling and vision.
How do I call it?
Through the standard chat completions endpoint at https://api.aimlapi.com/v1/chat/completions with the model id unbiased/pareto-26.10-preview. Any OpenAI-compatible client works, point the base URL at AI/ML API and use your AI/ML API key.
What is the Pareto 26.10 Preview model ID?
Use unbiased/pareto-26.10-preview as the model parameter. Requests go to the OpenAI-compatible endpoint with your AI/ML API key.
What is the context window of Pareto 26.10 Preview?
Pareto 26.10 Preview accepts up to 1,048,576 tokens of context and returns up to 131,072 output tokens.