nex-agi
Nex AGI: Nex-N2.5-Pro
nex-agi/nex-n2.5-pro
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
Nex-N2.5-Pro (nex-agi/nex-n2.5-pro) is the mid-size multimodal entry in Nex-AGI's Nex-N2.5 agentic family, accepting text and image input and producing text. The provider positions it for long-horizon agentic work: operating computers and browsers and executing programs with visual self-correction, building on the Nex-N2 line. The OpenRouter catalog lists a 262,144-token context, tool calling, structured outputs, and adjustable reasoning effort. Vendor-published benchmark tables cover coding, agentic, and computer-use tasks, though several rely on Nex-AGI's own harnesses. Card metadata lists an Apache-2.0 license.[1][2]
Strengths
- Provider-designed for long-horizon agentic tasks: computer and browser operation plus autonomous program execution, using vision as a feedback interface for self-correction rather than input only.[2]
- Broad listed serving surface: 262,144-token context, tools and tool_choice, structured outputs, logprobs, and configurable reasoning effort modes (none, medium, high).[1]
- The card lists Apache-2.0 licensing, states the family weights are being released open source, and documents a concrete single-node self-hosting path on 8xH100 via a provided sglang-based image.[2]
Limitations
- All performance figures are vendor-reported; several suites use Nex-AGI's own NexAU and NexCUA harnesses (NexCUA not yet open-sourced), so results need independent replication and validation on your own workloads.[2]
- Sources disagree on the default reasoning effort: the OpenRouter catalog lists high while the model documentation describes medium (adaptive) as the default; knowledge cutoff is not reported anywhere in the pack.[1][2]
- Evidence gaps: no independent evaluations, latency, throughput, or safety documentation were supplied; published guidance is catalog- and vendor-card-only.[1][2]
Where to use it
- Agentic coding pipelines that could benefit from a visual feedback loop (exploring codebases, multi-file edits, running and testing programs), per the provider's positioning; pilot on your own repositories first.[1][2]
- Browser- and desktop-automation agents where a long context and tool calling are useful; treat vendor computer-use scores as indicative only and validate reliability before production.[1][2]
- Multimodal tasks such as GUI grounding or document understanding; the vendor's multimodal tables are suggestive but self-evaluated, so confirm accuracy on representative inputs.[2]
Sources
- [1] OpenRouter catalog · Nex AGI: Nex-N2.5-ProSources checked: Nov 13, 2026
- [2] Model documentation · Nex AGI: Nex-N2.5-ProSources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
Nex-N2.5 is an agentic model built to turn goals into working, verified outcomes. Its core strength is agentic coding within a visual feedback loop: it can explore codebases, implement multi-file...
Specifications
- Model ID
- nex-agi/nex-n2.5-pro
- Canonical ID
- nex-agi/nex-n2.5-pro-20260907
- Creator
- nex-agi
- Added to OpenRouter
- 08.09.2026
- Context length
- 262,144
- Modality
- text+image->text
- Input modalities
- text, image
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $0.075 / out $0.250
- Knowledge cutoff
- —
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | No measurement | reasoning: optional · default high · efforts: high, medium, none · parameters: include_reasoning, reasoning, reasoning_effort |
| Coding | No measurement | — |
| Agents | No measurement | parameters: structured_outputs, tool_choice, tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | reasoning: optional · default high · efforts: high, medium, none · parameters: include_reasoning, reasoning, reasoning_effort |
| Document analysis | No measurement | — |
| Vision | No measurement | input: image |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | — |
| Embeddings | No measurement | — |
Evaluations
No evaluations are supplied for this model.
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| Nex AGIfp8 | 262,144 | $0.075 | $0.250 | — | — | 100% |
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.