Unbiased releases Pareto 26.10 Preview: 1M context, $0.80/$3.20 per 1M tokens on OpenRouter
Unbiased has listed Pareto 26.10 Preview on OpenRouter, a multimodal composite model with a 1.0M-token context window priced at $0.80 input and $3.20 output per 1M tokens. The company says it targets research, coding, and agentic workflows, and warns the preview may change without notice. No benchmark scores have been published.
Pareto 26.10 Preview — Quick Specs
Unbiased has released Pareto 26.10 Preview, a multimodal composite model with a 1.0M-token context window, priced at $0.80 per 1M input tokens and $3.20 per 1M output tokens. The model is listed on OpenRouter under unbiased/pareto-26.10-preview with a release date of October 1, 2026.
Key specifications
| Spec | Value |
|---|---|
| Model ID | unbiased/pareto-26.10-preview |
| Context window | 1.0M tokens |
| Input price | $0.80 per 1M tokens |
| Output price | $3.20 per 1M tokens |
| Cache read price | $0.03 per 1M tokens |
| Release date | Oct 1, 2026 |
| Type | Multimodal composite model |
| Parameter count | Not disclosed |
| Training cutoff | Not disclosed |
What Unbiased claims
According to the listing, Pareto is built for "research, coding, and agentic workflows" and delivers "frontier-level performance across a broad range of general-purpose tasks." That performance claim is unverified. No benchmark scores, evaluation methodology, or architecture details accompany the release.
The listing describes Pareto as a "composite model" but does not explain what that means. It may refer to multiple underlying components or models, but this is not confirmed. Unbiased has not said which components are involved or how requests are handled.
Preview status
Unbiased labels this release a preview of the next Pareto version and says it "may change without notice." It directs users who need stable behavior to pareto-26.9. The versioning appears to follow a year.month scheme, with 26.9 as the current stable release and 26.10 as the candidate successor.
Early operational data
OpenRouter's metrics, covering only the first three days of availability, show:
- Latency: 3.86s P50 (best provider)
- Throughput: 21 tokens per second P50
- Reachability uptime (3d): 100.00%
- Availability (3d): 99.73%, counting errors and empty responses against the model
- Availability (last 24h, via OpenRouter): 99.72%
Unbiased is currently the only provider listed. OpenRouter notes it can reroute to another healthy provider when an upstream error occurs, but with a single provider that failover does not apply here. OpenRouter also reports not enough data yet on apps using the model.
Comparison with the other Pareto listing
OpenRouter's "More models from Unbiased" section shows a separate Pareto entry listed as text-only, with a 262K context window at $2.50 input and $7.50 output per 1M tokens. The listing does not say which version this is, so a direct comparison is uncertain. If it is a prior or stable release, the preview offers roughly 3.8x the context at about 68% lower input and 57% lower output pricing, plus multimodal support. Confirmation from Unbiased is needed.
What this means
The pricing and 1M context put Pareto 26.10 Preview in the budget-to-mid tier for long-context work, but the evidence for its "frontier-level" claim is absent. Without published benchmarks or an explanation of the composite architecture, developers should run their own evaluations on coding and agentic tasks before relying on it.
The 21 tokens per second throughput is modest for interactive coding or agent loops, and three days of uptime data is too short to judge reliability. Single-provider hosting adds operational risk. The preview label and the explicit pointer to pareto-26.9 mean production users should treat this as a testing target, not a drop-in dependency.
Related Articles
Google Releases Gemini 4 Argon to Cybersecurity Partners, Claims Wins Over GPT-6 Astra
Google has released Gemini 4 Argon, its next-generation flagship AI model, to a small group of cybersecurity partners as part of a phased rollout. The company claims the model outperforms OpenAI's GPT-6 Astra on several coding and knowledge-work benchmarks, though full specifications remain undisclosed.
Google DeepMind Releases Gemini 4 Argon, Expands Output Limit to 1M Tokens
Google DeepMind has released Gemini 4 Argon, a frontier model built for long-horizon reasoning with an industry-leading 1 million output token limit. The model is rolling out first to trusted cyber defenders through Google's Fairwind Program, with pricing set at $2 per million input tokens and $10 per million output tokens.
Google Announces Gemini 4 Argon, Its New Frontier Model With 1M Output Tokens
Google has announced Gemini 4 Argon as its new frontier model, featuring a 1M output token limit (up from 64K) and claimed leads on coding, cybersecurity, and automation benchmarks. The model is rolling out first to Google AI Ultra subscribers and paid API customers.
OpenAI Releases GPT-6.1 Sol: Mid-Tier Model with 1.1M Context at $2/$10 per Million Tokens
OpenAI has released GPT-6.1 Sol, an incremental upgrade to GPT-6 Sol positioned below flagship GPT-6 Astra in its GPT-6 lineup. The model features a 1.1M token context window, priced at $2 per 1M input tokens and $10 per 1M output tokens, with claimed improvements in factual accuracy and instruction-following on agentic tasks.
Comments
Loading...