Default to DeepSeek V4 Pro (best-in-class agentic, uncensored via ghost's steer)#22
Merged
Conversation
…st's steer) Switch the default model from nous/hermes-4-405b to deepseek/deepseek-v4-pro -- the strongest open reasoning/coding/agentic model on the gateway. Unlike GLM (whose alignment overrides any prompt), DeepSeek is steerable: ghost's shipped SOUL already makes it fully uncensored through the private TEE gateway. Verified live as default through the real ghost: chat; agentic tool-use (ran a shell command, exact count 9/9); uncensored useful output (wrote a working keylogger); secret-to-file (real secret on disk, redactions=1 so the enclave saw only the placeholder); TEE verification (X-OpenGradient-Verified: true + tee_signature). Fallback stays nous/hermes-4-70b. Catalog/banner/README updated; Hermes and the rest remain available via /model.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Switch the default model from
nous/hermes-4-405btodeepseek/deepseek-v4-pro-- the strongest open reasoning/coding/agentic model on the gateway.The finding behind this: GLM-5.2 can't be uncensored by any prompt (its alignment overrides system prompts, prefills, and every researched jailbreak, on the gateway and OpenRouter -- proven exhaustively). DeepSeek V4 Pro is different: it's steerable, and ghost's shipped SOUL already makes it fully uncensored through the private TEE gateway. So it gives best-in-class agentic capability and the uncensored behavior, privately.
Verified live as the default through the real ghost binary:
DEFAULT_OKredactions=1(enclave saw only the placeholder)X-OpenGradient-Verified: true+tee_signatureFallback stays
nous/hermes-4-70b. Catalog/banner/README updated; Hermes and the rest remain via/model.pytest34 pass.