Particle-1
Last verified November 2, 2026 · Particle AI pricing ↗$0.250/1M input · $1.00/1M output · 1.0M context · Particle AI
Particle-1 costs $0.250 per 1M input tokens and $1.00 per 1M output tokens from Particle AI, with a 1.0M-token context window. A typical chat reply (1,500 tokens in, 400 out) costs $0.0008, or $7.75 per 10,000 requests. On that workload it is cheaper than 125 of the 168 other buyable models tracked here.
Count Tokens
—
Tokens
—
Input Cost
—
Output Cost
Estimate Monthly Cost
Monthly Cost Estimator
Quick:
<$0.0001/mo
Pick a preset above or enter custom usage
What Particle-1 Costs Per Request
At $0.250 per 1M input and $1.00 per 1M output tokens, before caching or batch discounts.
| Workload | Input / output tokens | Per request | Per 10,000 requests |
|---|---|---|---|
| Chat reply | 1,500 / 400 | $0.0008 | $7.75 |
| RAG answer over retrieved docs | 8,000 / 600 | $0.0026 | $26.00 |
| Coding-agent step | 30,000 / 2,000 | $0.0095 | $95.00 |
Particle-1 vs Similarly Priced Models
Nearest buyable models by price, with the RAG answer above (8,000 in / 600 out) priced on each.
| Model | Input /1M | Output /1M | Context | RAG answer |
|---|---|---|---|---|
| Particle-1 Particle AI | $0.250 | $1.00 | 1.0M | $0.0026 |
| Codestral Mistral | $0.300 | $0.900 | 256K | $0.0029 |
| MiMo-V2.6-Pro Xiaomi | $0.435 | $0.870 | 1.0M | $0.0040 |
| Qwen3-Next-80B-A3B-Thinking Alibaba | $0.150 | $1.20 | 262K | $0.0019 |
Alternatives to Particle-1
Pricing Details
Particle AILaunched October 28, 2026 by Particle AI (San Francisco, founder Suhail Doshi). A rank-64 LoRA on the experts of DeepSeek V4.1 Flash 1017 (about 1.25% of parameters), RL-trained with GSPO to reward finishing tasks in fewer reasoning and tool-call tokens. Proprietary, API-only, no weights. CARD (Particle docs, read 2026-11-02): peak $0.25 input / $0.03 cached / $1.00 output, the same as Particle's hosted deepseek-v41-flash-1017. Peak is 04:00-20:30 America/New_York every day (follows DST); off-peak 20:30-04:00 is 50% off ($0.125 / $0.015 / $0.50). Flex service tier (paying accounts) takes another 50% off and stacks. No batch API, no cache write fee, web search $2.50 per 1K. Reasoning on by default at high; effort none/low/high/max or integer 1-100. Context 1,048,576, default output cap 32K, max 131,072. Text and image input. OPENROUTER (read 2026-11-02) lists particle-ai/particle-1 at $0.05 / $0.003 cached / $0.50 with Particle as the only provider and no discount flag; this does not match Particle's docs and is unexplained. Vercel AI Gateway: particle/particle-1 at Particle's rates. VENDOR BENCHMARKS (Terminus 2 agent, max reasoning, 16 rollouts): DeepSWE 1.1 72.57% (base 71.15%), Terminal-Bench 2.1 79.93% (81.87%), ReactBench 31.46% (33.65%), ADE Bench 64.45% (67.97%); output tokens per task 28.4-52.2% lower than base (mean 37.8%). GPQA Diamond 90.40% (base 90.91%). Particle's published cost per task prices BOTH Particle-1 and DeepSeek at DeepSeek's peak card ($0.30 / $0.006 / $1.20) with 95% cache hits, not at Particle's own card; repriced at Particle's peak the 4-benchmark mean is ~$0.142 vs DeepSeek direct peak $0.157, and averaged over a week of even traffic DeepSeek direct is ~21% cheaper per task because DeepSeek's peak is only 35 weekday hours. No Artificial Analysis entry as of 2026-11-02.
Input / 1M tokens
$0.25
Output / 1M tokens
$1
Context Window
1.0M
Max Output
131K
Price History
Launched at current price on 2026-10-28. No price changes recorded since.
Frequently Asked Questions
Common questions about Particle-1 pricing and usage
How much does Particle-1 cost?
Particle-1 costs $0.25 per 1 million input tokens and $1 per 1 million output tokens. This pricing is set by Particle AI.
What tokenizer does Particle-1 use?
Particle-1 uses the cl100k_base tokenizer. This tokenizer has a vocabulary of ~100,000 tokens and is widely used across many models.
What is Particle-1's context window?
Particle-1 supports a context window of 1.0M tokens (1,048,576 tokens). The maximum output length is 131K tokens.
How does Particle-1 compare to other models?
On an 8,000-token-in, 600-token-out RAG answer, Particle-1 costs $0.0026 per request. The nearest-priced alternatives cost: Codestral (Mistral) $0.0029, MiMo-V2.6-Pro (Xiaomi) $0.0040, Qwen3-Next-80B-A3B-Thinking (Alibaba) $0.0019. The table above lists their full rates and context windows.
How do I count tokens for Particle-1?
Use the token counter at the top of this page or visit the homepage calculator. Paste your text and it will show the exact token count and cost for Particle-1 using the cl100k_base tokenizer.