Particle AI: Particle-1

Image input Text input Text output Free Option
Author's Description

Particle-1 is a token-efficient agentic coding model from Particle AI, post-trained on DeepSeek V4.1 Flash. It is suited for software engineering work such as terminal workflows, codebase changes, and data...

Key Specifications
Cost
$$$$
Context
1M
Released
Oct 28, 2026
Speed
★★★★
Ability
★★★★
Reliability
★★★★★
Supported Parameters

This model supports the following parameters:

Frequency Penalty Presence Penalty Min P Top P Max Tokens Seed Tools Include Reasoning Logprobs Stop Temperature Top Logprobs Response Format Tool Choice Structured Outputs Reasoning
Features

This model supports the following features:

Reasoning Tools Structured Outputs Response Format
Performance Summary

Particle-1 from Particle AI delivers strong overall capabilities, pairing top-tier execution speeds with competitive pricing across comprehensive evaluations. Operating in the 73rd percentile for speed while maintaining balanced operational costs, the model also demonstrates exceptional technical reliability, recording a perfect 100% success rate with zero system or API failures. Performance across benchmark categories is anchored by exceptional analytical and technical strengths. Most notably, Particle-1 achieved a perfect 100% accuracy in complex reasoning, distinguishing itself as the most accurate model at its price point and within its speed class. It also excels in programming, securing a 95.0% accuracy score (87th percentile) in coding assessments, alongside strong mathematical proficiency at 94.0%. The model's primary relative weakness appears in instruction following, where it attained a lower 78.0% accuracy when navigating complex, multi-layered constraints. Nevertheless, bolstered by its massive 1M-token context window and specialized post-training, Particle-1 offers an outstanding balance of speed, affordability, and elite problem-solving, making it particularly well-suited for demanding software engineering tasks and data-heavy agentic workflows.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.06
Completion $0.5
Input Cache Read $0.003

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Particle AI
Particle AI | particle-ai/particle-1-20261027 1M $0.06 / 1M tokens $0.5 / 1M tokens
Particle AI
Particle AI | particle-ai/particle-1-20261027 1M $0.06 / 1M tokens $0.5 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Strategy Free Executions Accuracy Cost Duration