Choose the models AI2Apps should use when a caller asks for a task class or dedicated capability without naming a model. Empty slots inherit the API default.
A shared contract for Agents, Apps, Services, and system automation.
Only compatible models from the current local, Fusion, and enabled cloud catalog are offered.
{{ t('models.manager.description') }}
Combine a local generator, a reviewer, and an optional external resolver.
Defaults for final review, 1K checkpoints, and <think> auditing. Chat sessions may override these.
Default optimization policy for the generator model.
Default optimization policy for the reviewer model.
External assistance requires the reviewer itself to be local.
Called at most once for reviewer escalation, low confidence, or patch failure. The provider receives the conversation, draft, and local review. Limit: 384 output tokens and 30 seconds. On failure, Fusion falls back to the local review blueprint.
Install a local model before creating a Fusion model.
The key remains local. AI2Apps requests the Provider's Models API and stores only the returned model metadata.
{{ t('models.downloader.description') }}
{{ t('models.browse.loading') }}
{{ t('models.browse.load_prompt') }}
|
{{ t('models.browse.tight_fit') }}
{{ t('models.browse.high_oom_risk') }}
|
{{ t('models.browse.tight_fit') }}
{{ t('models.browse.high_oom_risk') }}
|
||||
{{ t('models.browse.searching') }}
{{ t('models.browse.search_prompt') }}
{{ t('models.browse.no_results') }}
|
{{ t('models.browse.tight_fit') }}
{{ t('models.browse.high_oom_risk') }}
|
{{ t('models.browse.tight_fit') }}
{{ t('models.browse.high_oom_risk') }}
|
{{ t('models.browse.loading') }}
{{ t('models.browse.load_prompt') }}
{{ t('models.browse.no_trending') }}
{{ t('models.browse.no_popular') }}
{{ t('models.browse.searching') }}
{{ t('models.browse.search_prompt') }}
{{ t('models.browse.no_results') }}
{{ t('models.oq.description') }}
{{ t('models.oq.enhanced_help') }}
{{ t('models.oq.imatrix_reuse_cache_help') }}
{{ t('models.oq.imatrix_strict_help') }}
Excludes vision encoder weights. Output is a text-only model (~2-3% smaller).
{{ t('models.oq.preserve_mtp_help') }}
{{ t('models.oq.preserve_mtp_unavailable') }}
{{ t('models.oq.combine_assistant_help') }}
{{ t('models.oq.combine_mtp_donor') }}{{ t('models.oq.combine_mtp_donor_help') }}
{{ t('models.oq.dtype_help') }}
Quantization should not be exclusive to any particular inference server.
oQ produces standard mlx-lm models that work everywhere — oMLX, mlx-lm, LM Studio, and any app that supports MLX safetensors format.
No custom loader required.
The quantizer streams tensors from safetensors, measures layer sensitivity, and writes a byte-budgeted mixed-precision checkpoint. Standard oQ focuses on layer-level sensitivity and model-aware protection rules. oQe keeps that same oQ plan and adds imatrix-weighted affine quantization.
oQe uses the same oQ sensitivity planner, then adds an importance-matrix calibration step. This is explicitly borrowed from the llama.cpp imatrix idea: collect activation energy from representative prompts, then use that information so quantization spends less error on the input channels that matter most.
| Path | Bit allocation | Affine quantization | Best use |
|---|---|---|---|
| oQ | Layer-sensitivity mixed precision | Standard min/max affine per group | Fast, deterministic conversion with strong baseline quality |
| oQe | Same oQ plan | imatrix-weighted clipping/search per group | Better low-bit retention, especially for MoE and activation-skewed layers |
If a tensor has no matching imatrix entry, the default behavior is to fall back to standard oQ for that tensor. Enable strict coverage to fail instead.
{{ t('models.detail.loading') }}
{{ t('models.detail.no_model_card') }}
{{ t('models.detail.no_files') }}
{{ t('models.detail.no_tags') }}