Outils open source et modèles open-weight, adaptés à l’appareil.
selbsai est configuré en fonction des limites de mémoire, de thermique et de latence. Le configurateur raisonne donc en classes de modèles plutôt qu’en nom marketing figé. La pile combine logiciels open source classiques et familles de modèles ouvertes, puis ajuste le déploiement au tier matériel, aux langues, au workload et aux contraintes de licence.
7B-8B class
Open-weight 7B-8B instruct stack
Fast local assistance for drafting, document Q&A, responsive chat and day-to-day offline copilots.
- RAM minimale
- 16 GB
- Vitesse cible
- 15-40 t/s
7B-13B class
Open-weight 7B-13B reasoning and retrieval stack
Balanced local reasoning for research, document-heavy workflows, coding support and assistant-style agents.
- RAM minimale
- 32 GB
- Vitesse cible
- 12-30 t/s
13B-30B class
Open-weight 13B-30B advanced local model stack
High-memory local intelligence for larger models, orchestration, coding agents and heavier secure workloads.
- RAM minimale
- 64 GB
- Vitesse cible
- 8-18 t/s
Familles représentatives que nous surveillons réellement.
Ces familles servent à calibrer les tiers actuels de selbsai. Les modèles réellement livrés peuvent changer à mesure que de nouvelles sorties ouvertes prouvent leur valeur sur des benchmarks indépendants.
Llama 3.1 8B Instruct
Meta Llama · Meta
Our compact baseline for fast local assistants, offline drafting and lighter retrieval on personal desktop nodes.
Ministral 3 14B Instruct
Mistral / Ministral · Mistral
A strong local step-up for multilingual work, longer-context tasks and richer instruction-following on balanced desktop systems.
Gemma 4 with MTP drafters
Google Gemma · Google DeepMind
A model family we track for responsive local chat, coding assistants and agentic workflows because multi-token prediction drafters can accelerate local inference while the main model verifies outputs.
Qwen3 30B A3B / 32B class
Qwen · Qwen
One of the main higher-capability open families we watch for larger reasoning, coding and multilingual deployments on high-memory local nodes.
Artificial Analysis
Independent model pages with direct comparisons across intelligence, speed, price, context window and methodology notes.
Hugging Face Open LLM Leaderboard
A widely used open-model benchmark hub for comparing community and lab releases across standard eval suites.
Arena Leaderboard
Useful for broad human-preference comparisons and keeping an eye on how major open releases stack up in live arena-style evaluation.
Ce qui est réellement installé
Nous ne promettons pas qu’une unité Standard, Professional ou Elite expédie toujours le même modèle nommé. Un vendeur sérieux de matériel IA local doit choisir la bonne pile open-weight pour le besoin réel et mettre cette recommandation à jour quand l’écosystème progresse.
- Le tier détermine l’échelle de modèle possible, le plancher mémoire et l’enveloppe de latence.
- Les presets de workflows déterminent l’outillage autour du modèle de base : code, revue documentaire, retrieval, opérations ou audit.
- Le choix final dépend des langues, de la sensibilité des données, des contraintes de licence et de l’équilibre recherché entre vitesse, profondeur et multimodalité.
- Les positions sur les benchmarks évoluent. Cette page pointe donc vers des références vivantes de tiers au lieu de figer des promesses périmées.
Hugging Face scale is useful only after filtering.
The value for customers is not simply that open models exist. The value is that selbsai turns a fast-moving model ecosystem into a controlled local setup with documented choices, workload fit, and a clear update channel.
Source reputation
Publisher history, release notes, model-card quality, community usage, and maintenance signals are reviewed before a model is treated as a provisioning candidate.
License and usage fit
The configurator now captures whether the customer wants permissive-only, commercial-ready, or restricted-model avoidance before final model selection.
Safe format preference
Where supported, selbsai prefers formats and runtimes with clearer supply-chain posture, including Safetensors, GGUF, MLX packages, and established local runtimes.
Hardware match
The selected model class is checked against RAM, VRAM, thermal budget, storage, context length, and the customer's target workloads.
What the customer should know about the installed stack.
- Model family, exact source repository, publisher, model-card link, and release reference.
- Runtime path, file format, quantization level, checksum or verification reference where available.
- License posture, intended use, known limitations, language fit, and benchmark references.
- Selected update policy: stable, balanced, or fast track.
OCR and document extraction
For invoice, receipt and document-heavy presets we pair the language model with open OCR and document-understanding tooling rather than relying on the base LLM alone.
Retrieval and reranking
Search-heavy presets use additional embedding and reranking components so large local indexes stay usable at real-world scale.
Workloads prédéfinis
Software coding
Local help for private repositories.
Explain code, draft tests, review snippets, write scripts, and search repository notes without sending proprietary source to cloud tools.
- Repo-aware Q&A
- Test and script drafts
- Error explanation
Documents and writing
Draft, rewrite, summarize, extract.
Create letters, policies, proposals, memos, summaries, and structured extracts from files that should stay inside your office.
- PDF & DOCX ingestion
- Memo and report drafts
- Tables and summaries
Email and personal assistant
Inbox work without inbox exposure.
Draft replies, sort messages, extract tasks, prepare agendas, and turn notes into follow-ups from approved local exports.
- Reply drafts
- Action extraction
- Meeting follow-ups
Research desk
Turns reading piles into briefings.
Compare sources, summarize PDFs, answer questions with citations, and prepare decision notes from local research folders.
- Citation-aware Q&A
- Long-context search
- Briefing notes
Document review
Find clauses, risks, gaps, and dates.
Review contracts, policies, case files, leases, and due-diligence packs for obligations, inconsistencies, and missing attachments.
- Clause search
- Obligation extraction
- Risk and gap lists
Sales assistant
Prepare better conversations faster.
Draft outreach, summarize accounts, prepare call notes, handle objections, and build proposals from approved sales material.
- Proposal drafts
- Call preparation
- CRM-style summaries
Compliance management
Policies and evidence, searchable locally.
Answer audit questions, compare obligations, identify missing evidence, and prepare control summaries from internal policy folders.
- Policy Q&A
- Evidence checklists
- Audit response drafts
Warehouse management
Operations support from local records.
Search SOPs, summarize shift notes, prepare supplier messages, and answer operational questions from warehouse documentation.
- SOP search
- Shift note summaries
- Supplier message drafts
Inventory management
Stock lists, reorder issues, and reports.
Review stock exports, flag reorder risks, summarize item movements, and prepare plain-language inventory reports.
- CSV and table review
- Reorder flags
- Inventory summaries
Company knowledge base
Ask your manuals, folders, and notes.
Build a local question-answer layer over manuals, procedures, project folders, email exports, and internal documentation.
- Local vector index
- Folder Q&A
- Source-grounded answers