AllModels

About AllModels

AllModels is a model router that connects to abliterated language models across multiple providers through a single API. It includes automatic fallbacks, modality-aware selection, and a dashboard for comparing availability and token prices. The tool is built for developers who need to route requests to abliterated models without managing separate provider integrations.

Review

AllModels launched recently as a specialized routing layer for abliterated models - models that have undergone a safety-refusal removal process. The service consolidates provider access, letting developers point to one endpoint instead of juggling multiple API keys and libraries. The router handles provider selection and failover, though the feature set is still early and some details remain undefined.

Key Features

• Unified API for abliterated models. One endpoint routes requests to abliterated models from multiple providers, removing the need for separate integrations.
• Automatic fallbacks. If a primary provider fails or times out, the router shifts the request to an alternative provider. Response metadata indicating which provider or model handled the request is a requested feature, not yet confirmed as shipped.
• Modality-aware selection. The router factors input modality into its selection logic, aiming to avoid routing decisions that work for text but miss on other input types.
• Provider comparison dashboard. A single interface displays availability, latency, and token pricing across supported providers.

Pricing and Value

AllModels has not published a pricing page or defined its pricing model at this stage. The comparison dashboard shows token prices across providers, but whether the router adds its own markup, charges a flat fee, or operates on a different structure is not yet stated. Developers evaluating the tool should confirm pricing directly before integrating into production workloads.

Pros

• Reduces the number of provider-specific integrations a developer needs to maintain for abliterated models.
• Fallback mechanism keeps requests flowing when a single provider experiences an outage.
• Modality-aware routing addresses a real failure mode where text-optimized routers mishandle other input types.
• Comparison view puts provider availability and token costs side by side in one place.
• Team is actively responding to early feedback on feature requests like fallback metadata.

Cons

• Pricing is not yet defined, making cost projections impossible for production planning.
• The router is not well suited for teams that don't use abliterated models or need fine-grained control over which provider handles each request.
• Fallback metadata - whether a different provider or model handled a request - is not yet confirmed as available in responses, limiting debugging visibility.

AllModels fits developers who already work with abliterated models and want to simplify provider management through a single API. The modality-aware routing and automatic failover address concrete operational concerns, but the undefined pricing and missing fallback metadata mean early adopters should test thoroughly before committing. Teams that don't require abliterated models or need direct provider control will find the tool less relevant.