HyperNova Models by Multiverse Computing
Multiverse Computing releases · 1 in the last 12 monthsChangelog →
1 model2026Up to 131k ctx
Last refreshed 2026-10-05. Next refresh: weekly.
Details
ResearcherMultiverse Computing
LicenseApache 2.0OSI-approved
Commercial useCommercial use: permitted
Models1
Released2026
Max context131k
Capabilities
ReasoningAll models
JSON / Tool useAll models
Structured OutputsAll models
About
HyperNova is Multiverse Computing's line of CompactifAI-compressed open-weight models built on OpenAI's gpt-oss-120b. Tensor-network compression roughly halves the size (~60B total / 4.8B active, ~32 GB) to fit on one 40 GB GPU, and knowledge distillation recovers lost capability. Releases: HyperNova 60B (Jan 2026), 2602 (tool calling) and 2605 (coding/general gains), all Apache 2.0 on Hugging Face. The first-party CompactifAI API deprecated hypernova-60b on 2026-09-09 in favour of Carina 60B.
Decision facts
- Best fit
- reasoningJSON / Tool usestructured outputs
- Capability starting point
- HyperNova 60B with 131k context and reasoning, JSON / Tool use, and structured outputs
- Lowest tracked input
- Not tracked
- Closest related family
- 360Zhinao
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
1 in view
HyperNova 60BCurrent
Use when the workload needs 131k context, reasoning, and JSON / Tool use.
2026-01131k contextreasoningJSON / Tool use
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| HyperNova 60B | Use when the workload needs 131k context, reasoning, and JSON / Tool use. | 2026-01 | 131k contextreasoningJSON / Tool use | Current |
Release Timeline
1 release group2026-01
1 current
HyperNova 60B
Current131k contextreasoningJSON / Tool use
Specifications(1 models)
| Model | Released | Context | Parameters | Reasoning | JSON / Tool use | Structured Outputs |
|---|---|---|---|---|---|---|
| HyperNova 60B | 2026-01 | 131k | 60B (4.8B active) | Yes | Yes | Yes |