better-model-provider
DeepSeek Harness settings extension for non-destructive per-model reasoning-effort levels, input modalities, and overrides.
Install
$ dsh plugin --profile web add better-model-providerPlugin Overview & Capabilities
AI-assisted organization based on the public repository snapshot. The content must be grounded in source evidence and does not replace compatibility or security verification.
Key Capabilities
- Adds a 'Model capabilities' settings section to DeepSeek Harness web interface
- Configures custom reasoning-effort levels and wire spellings per model
- Toggles input modalities such as image/vision support for multimodal models
- Applies sparse modelOverrides on official catalog routes without replacing upstream catalog definitions
- Edits model declarations directly on custom declared routes
- Provides quick recovery mechanisms like resetting to official defaults and cleaning leftover overrides
Useful For
- Enabling vision/multimodal capabilities on specific OpenAI-compatible provider models
- Customizing reasoning effort parameters for specialized reasoning models
- Adjusting context windows and token limits on catalog routes while keeping upstream route tracking intact
- Fine-tuning model parameter overrides without editing raw YAML configuration files
Who It Fits
- DeepSeek Harness web users working with multiple OpenAI-compatible or reasoning models
- Developers needing granular control over LLM request modalities and reasoning effort wires
- Users who want to avoid manual YAML editing for per-model catalog overrides
Documented Limitations
- Requires API keys and base provider routes to be configured first on the official Models page
- Does not configure dedicated-adapter apps such as built-in DeepSeek or OpenAI Codex that declare settings on their own pages
- Write validation requires explicit removal of leftover overrides if a route's whole model list was already rewritten on the official Models page
DSH Compatibility
Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.
Security Signals
Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.
package.json declares 0 runtime, 16 development, 6 peer, and 0 optional dependencies.
package.json declares DSH bundle metadata.
GitHub reports the repository license as MIT.
A root package.json was captured and can be inspected by the registry.
Public GitHub source metadata is available for this registry snapshot.
Source & Registry Notes
Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.
- Source repository
- sanshanya/better-model-provider
- Registry source
- GitHub · dsh-plugin topic
- Registry classification
- Plugin
- Source checked
- b2771e3 · 2026-10-15
This project is independently indexed from public source information. DSH Plugin is not affiliated with DeepSeek or the plugin author. Always check the author repository before installation.
Repository Activity
- GitHub stars
- 13GitHub stars
- Forks
- 0
- Open issues
- 1
- Last commit
- 2026-10-09
- Last release
- 2026-10-09
Related DSH Plugins
Ranked by overlapping capabilities, use cases, plugin type, categories, and DSH profile.
Infinite canvas UI plugin for DeepSeek Harness that turns branching LLM conversations into an editable thought graph.
View plugin →Context visualization, inspection, and lifecycle management dashboard and /context command for DeepSeek Harness.
View plugin →Generative UI plugin for DeepSeek Harness that renders interactive cards, charts, forms, and panels inline from dsh-ui code fences.
View plugin →Agent-native trading terminal and multi-market plugin bundle for DeepSeek Harness covering Crypto, US, CN, and HK markets.
View plugin →