Back to Models
AionLabs: Aion-3.0
Aion Labstext7.8 / 10 Overall Rating
Developed by AionLabs on the GLM base architecture, Aion-3.0 is a multi-model system optimized for roleplay and narrative generation. It pairs a 131K context window with an unusually high 32K max output token limit at $3.00/$6.00 per million tokens. Unlike general-purpose benchmarks like GPT-4o and Claude 3.5 Sonnet, it focuses primarily on collaborative creative writing rather than technical reasoning or coding.
Context Window
131K
Knowledge Cutoff
2025
Max Output
16K
License
Proprietary
Where it Excels (Pros)
- 32K maximum output token window
- Specialized for narrative and roleplay
- Cheaper output pricing than Sonnet
Limitations (Cons)
- Unproven on technical coding tasks
- Higher input cost than GPT-4o
- Limited general enterprise utility
Benchmark Breakdown
Reasoning (MMLU)82%
Coding (HumanEval)78%
Mathematics (MATH)76%
Long-Context Retrieval82%
Pricing Matrix
Input cost / 1M tokens$3.00
Output cost / 1M tokens$6.00
Compare AionLabs: Aion-3.0 with other frontier models: