Pareto 26.10 Preview

Text input Image input Text output
Author's Description

Pareto is a multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of general-purpose tasks. This is a preview of the...

Key Specifications
Cost
$$$$
Context
1M
Parameters
26.1B (Rumoured)
Released
Oct 01, 2026
Speed
★★
Ability
★★★★
Reliability
★★★★★
Supported Parameters

This model supports the following parameters:

Top P Temperature Tool Choice Tools Max Tokens
Features

This model supports the following features:

Tools
Performance Summary

Pareto 26.10 Preview, a multimodal composite model from unbiased, demonstrates a balanced performance profile across various benchmarks. The model exhibits moderate speed performance, ranking in the 40th percentile, and offers competitive pricing, placing it in the 37th percentile. A standout feature is its exceptional reliability, achieving a 100% success rate across all evaluated benchmarks, indicating consistent and stable operation with minimal technical failures. In terms of specific capabilities, Pareto 26.10 Preview shows strong performance in Instruction Following and Reasoning, achieving 83.0% and 98.0% accuracy respectively, both ranking in the 87th percentile. This suggests a significant strength in understanding and executing complex directives, as well as in tackling intricate logical and abstract problems. Its performance in Mathematics (92.0% accuracy, 54th percentile) and Coding (92.0% accuracy, 66th percentile) is solid, indicating a broad general-purpose capability. While its mathematical understanding is robust, its instruction following and reasoning abilities appear to be its primary strengths, aligning with its description as a model built for research, coding, and agentic workflows.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.8
Completion $3.2
Input Cache Read $0.03

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Unbiased
Unbiased | unbiased/pareto-26.10-preview-20260929 1M $0.8 / 1M tokens $3.2 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Strategy Free Executions Accuracy Cost Duration
Other Models by unbiased