The Open Weights
LatestModelsLeaderboardsCompanies
Subscribe
The Open Weights

The daily record of open-source AI. New model releases, leaderboards, and what's coming next — written for people who ship.

Refreshed every 12 hours

Discover

  • Latest releases
  • New today
  • Trending models

Browse

  • All models
  • Companies
  • Categories
  • Leaderboards

About

  • About
  • Editorial policy
  • Hardware estimates
  • RSS feed
  • llms.txt
  • Newsletter

© 2026 The Open Weights. An independent publication.

PrivacyTermsSMSAggregated by Claude · curated by humans.
LatestOVHaiLLMv1.0-mini
OVHaiLLMText / LLM

OVHcloud enters AI race with Olala, a Mamba/MoE model

The French cloud provider's first language model is a compact hybrid architecture built with EU public funding.

Oct 30, 2026
NotableOther
Olala v1.0 mini 7A1B

OVHcloud, the French cloud and hosting company, has released its first language model. Called Olala v1.0 mini 7A1B, the model marks the provider's entry into building its own open-weight AI rather than simply hosting others' systems.

The design is unusual for a debut. Olala combines a Mamba-style state-space backbone with a mixture-of-experts layer, giving it 7 billion total parameters but only around 1 billion active per token. That approach aims to keep inference costs low while retaining the capacity of a larger network — a sensible priority for a company whose business is efficient infrastructure.

Why it matters

The release is notable less for raw benchmarks, which the record does not detail, than for who is behind it. Olala was built with EU public funding, part of a broader European push to develop sovereign AI capabilities that don't depend on US or Chinese labs.

  • Hybrid Mamba plus mixture-of-experts architecture
  • 7B total parameters, roughly 1B active
  • Listed for text generation and reasoning
  • Released under an "other" license, so terms warrant a close read

As a "mini" first version, Olala reads more like a statement of intent than a flagship. The real test will be whether OVHcloud follows with larger models and clearer licensing — and whether a European cloud provider can turn infrastructure scale into competitive open models.

Sources

  • OVHaiLLM/Olala-v1.0-mini-7A1B

    Hugging Face

    Visit
OlderIBM's Granite Switch Treats Models Like SoftwareIBM · Any-to-Any · 3 weeks agoNewerMicrosoft's Mage-VL-Cowork takes on computer-use tasksMicrosoft · Vision-Language · 3 weeks ago

Get the model

Hugging Face

Specs

Parameters7B · MoE
Active params1B active
Context window100K tokens
Size13.6 GB
PrecisionBF16
ArchitectureOlalaForCausalLM
LicenseOTHER
Downloads62.8K
Likes221

Can you run it?

Runs on a laptop — about 12.3 GB at 4-bit.

  • BF16 (as published)

    24 GB GPU (RTX 3090 / 4090) · 32 GB Mac

    21.8 GB
  • 8-bit

    16 GB GPU (RTX 4080 / 5070 Ti) · 24 GB Mac

    15.4 GB
  • 4-bit

    16 GB GPU (RTX 4080 / 5070 Ti) · 24 GB Mac

    12.3 GB
Your machine
GGUF builds
  • Mixture-of-experts: all 6.8B parameters must be resident even though only 1B are active per token — active count affects speed, not memory.
  • KV cache sized for 8,192 tokens of context; longer prompts need proportionally more.

Based on Hugging Face weights metadata. Assumes an 8K context and ~1 GB runtime overhead; actual needs vary. How we estimate


Modalities

Text / LLMReasoning

The Weekly Weights

Every open release that mattered, one email a week.

0 comments

No comments yet. Be the first to weigh in.

More in Text / LLM

All Text / LLM →
Aion-1.0-Instruct-2.3B
Microsoft/Text / LLMRuns anywhere

Microsoft's Aion 1.0 Instruct is a compact 2.3B chat model

A small, multilingual instruction-tuned model aimed at chat and summarization lands on Hugging Face.

Nov 18, 2026
SenseNova-Loop-9B
SenseTime/Vision-LanguageRuns on a laptop

SenseTime Releases Looped 9B Multimodal Model

The company's new SenseNova-Loop-9B pairs vision-language capabilities with a recurrent looped-transformer design.

Nov 17, 2026
Hunyuan Hy4-flash
Tencent/Vision-LanguageMulti-GPU server

Tencent releases Hunyuan Hy4-flash multimodal MoE

The lightweight, mixture-of-experts model adds vision and text instruction-following to Tencent's open Hunyuan lineup.

Nov 17, 2026