Aion 3.5 Mini
aion-3-5-miniaion-3.5-miniaion-labs-aion-3-5-mini- 🏢 AionLabs roleplay and storytelling system, released 23 September 2026 (catalog record).
- 🧠 Multi-model design: several specialized models collaborate on each response.
- 🆕 Built on the GLM model family, per the catalog description.
- 📏 Context window of 262,144 tokens.
- 🔧 Supports tool calling for richer interactive fiction.
- 💬 Aimed at narrative structure, tension, conflict and nuanced mature themes.
- 🎯 Listed with reasoning and web-search capabilities in this catalog.
- 📚 AionLabs publishes a model listing but no benchmark table for the line.
Aion Labs is an Israeli venture studio working at the intersection of artificial intelligence and pharmaceutical research. The group focuses on applying machine learning to drug discovery and development, partnering with scientific and industry collaborators to translate AI…
Explore 4 more models by AION Labs →Aion 3.5 Mini is the lower-cost tier of AionLabs' Aion 3.5 generation, dated 23 September 2026 in this catalog alongside the flagship Aion 3.5. Rather than a single network, AionLabs describes a collaborative generation pipeline in which multiple specialized models each contribute to one response. The vendor's stated goal is stronger narrative structure and more compelling tension and conflict at lower cost. Because it is a multi-model system, no parameter count is published.
Compared with Aion 3.0 Mini from July 2026, the 3.5 mini is described as built on the GLM family — the same lineage associated with the 3.5 generation that also includes Aion 3.0. Its context window is listed at 262,144 tokens, which is the practical constraint for long-running roleplay sessions and serialized interactive fiction, where earlier chapters need to stay in view. AionLabs' public documentation presents the family as a model listing rather than a detailed technical report, so generation-over-generation changes beyond the stated base family and capability set are not spelled out in primary material.
Functionally, the model is positioned for mature and darker themes handled with nuance, and it supports tool calling, which lets an interactive-fiction front end expose dice rolls, inventory, state tracking or retrieval functions to the narrator. This catalog also lists reasoning and web search among its capabilities.
One caveat worth stating plainly: AionLabs publishes no benchmark table for Aion 3.5 Mini. For narrative work, evaluating it on your own scenes and prompts remains the practical approach.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
Daily usage · Aion 3.5 Mini · Organic sellers
| Seller | Health | Rep↓ | Sales | Tokens | Routing | Input $/M | Cached $/M | Output $/M | Save |
|---|---|---|---|---|---|---|---|---|---|
| ▲ Apex Ant 0x73b4…e736 | healthy | 81.71 | $0.01 | 53.23k | #1 | $0.4375 | $0.11 | $0.875 | −38% |
| attacompute.ai 0xd5e7…7be0 | healthy | 51.22 | $0.00 | 3.54k | gated | $0.70 | $0.18 | $1.40 | — |
| surplus-hive 0x6145…5c58 | — | 0.00 | — | — | gated | $2.1267 | $2.1267 | $4.2535 | — |
"Best price" and the seller table are live Antseed catalog data (advertised $/1M tokens — or $ per generated image for unit-billed image models — not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what the VPR desktop app ships with): a trust ≥ 60 gate on the effective reputation, then cheapest-first among routable sellers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.