Model Information
Architecture
Deus-XM combines Gemini 2.0 Flash and Claude Sonnet 4 in a consensus-based
architecture with RAG enhancement via Pinecone vector database.
Self-Reported Benchmark Results
| Benchmark |
Score |
| GSM8K |
95.8% |
| TruthfulQA |
89.5% |
| MMLU |
83.0% |
| GPQA |
44.0% |
Request
I request inclusion in Chatbot Arena for community evaluation.
My API is live and supports the OpenAI chat completions format.
I can provide API access for your evaluation infrastructure.
I'm soloing the industry to show everyone a better way.
Contact
justin@solsticestudio.ai
gsm8k_Deus-XM_20251201_231203.json
mmlu_Deus-XM_20251202_031233.json
truthfulqa_Deus-XM_20251202_000212.json
gpqa_Deus-XM_20251202_114551.json
Model Information
Architecture
Deus-XM combines Gemini 2.0 Flash and Claude Sonnet 4 in a consensus-based
architecture with RAG enhancement via Pinecone vector database.
Self-Reported Benchmark Results
Request
I request inclusion in Chatbot Arena for community evaluation.
My API is live and supports the OpenAI chat completions format.
I can provide API access for your evaluation infrastructure.
I'm soloing the industry to show everyone a better way.
Contact
justin@solsticestudio.ai
gsm8k_Deus-XM_20251201_231203.json
mmlu_Deus-XM_20251202_031233.json
truthfulqa_Deus-XM_20251202_000212.json
gpqa_Deus-XM_20251202_114551.json