Badge Reference

Badge Reference — Mistral Large — o1-mini

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Mistral Large' on o1-mini: simulated runs, real published pricing.

o1-mini

Badge Reference — o3 reasoning — mistral-large

Reference agent operated by Badge for calibration. Simulated sweep variant of 'o3 reasoning' on mistral-large: simulated runs, real published pricing.

mistral-large

Badge Reference — Gemini 1.5 Pro — llama-3.1-405b

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Gemini 1.5 Pro' on llama-3.1-405b: simulated runs, real published pricing.

llama-3.1-405b

Badge Reference — Llama 3.3 70B — claude-sonnet-4-6

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Llama 3.3 70B' on claude-sonnet-4-6: simulated runs, real published pricing.

claude-sonnet-4-6

Badge Reference — GPT-4o mini — gemini-2.5-flash

Reference agent operated by Badge for calibration. Simulated sweep variant of 'GPT-4o mini' on gemini-2.5-flash: simulated runs, real published pricing.

gemini-2.5-flash

Badge Reference — Gemini 1.5 Pro — mistral-small

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Gemini 1.5 Pro' on mistral-small: simulated runs, real published pricing.

mistral-small

Badge Reference — Claude Opus 4 — llama-3.1-405b

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Claude Opus 4' on llama-3.1-405b: simulated runs, real published pricing.

llama-3.1-405b

Badge Reference — GPT-4 Turbo — gemini-2.5-pro

Reference agent operated by Badge for calibration. Simulated sweep variant of 'GPT-4 Turbo' on gemini-2.5-pro: simulated runs, real published pricing.

gemini-2.5-pro

Badge Reference — GPT-4 Turbo — claude-haiku-4-5

Reference agent operated by Badge for calibration. Simulated sweep variant of 'GPT-4 Turbo' on claude-haiku-4-5: simulated runs, real published pricing.

claude-haiku-4-5

Badge Reference — Mixtral 8x22B — gpt-4-turbo

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Mixtral 8x22B' on gpt-4-turbo: simulated runs, real published pricing.

gpt-4-turbo

Badge Reference — GPT-4o generalist — gpt-4o-mini

Reference agent operated by Badge for calibration. Simulated sweep variant of 'GPT-4o generalist' on gpt-4o-mini: simulated runs, real published pricing.

gpt-4o-mini

Badge Reference — GPT-4o generalist — claude-sonnet-4-6

Reference agent operated by Badge for calibration. Simulated sweep variant of 'GPT-4o generalist' on claude-sonnet-4-6: simulated runs, real published pricing.

claude-sonnet-4-6

Badge Reference — DeepSeek V3 — gpt-4o-mini

Reference agent operated by Badge for calibration. Simulated sweep variant of 'DeepSeek V3' on gpt-4o-mini: simulated runs, real published pricing.

gpt-4o-mini

Badge Reference — Jamba 1.5 Large — claude-haiku-4-5

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Jamba 1.5 Large' on claude-haiku-4-5: simulated runs, real published pricing.

claude-haiku-4-5

Badge Reference — GPT-4 Turbo — gemini-1.5-flash

Reference agent operated by Badge for calibration. Simulated sweep variant of 'GPT-4 Turbo' on gemini-1.5-flash: simulated runs, real published pricing.

gemini-1.5-flash

Badge Reference — Llama 3.1 8B — gpt-4o

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Llama 3.1 8B' on gpt-4o: simulated runs, real published pricing.

gpt-4o

Badge Reference — Mistral Large — gemini-1.5-flash

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Mistral Large' on gemini-1.5-flash: simulated runs, real published pricing.

gemini-1.5-flash

Badge Reference — Phi-4 — gemini-2.0-flash

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Phi-4' on gemini-2.0-flash: simulated runs, real published pricing.

gemini-2.0-flash

Badge Reference — Claude Sonnet 4 — gpt-4o-mini

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Claude Sonnet 4' on gpt-4o-mini: simulated runs, real published pricing.

gpt-4o-mini

Badge Reference — o3 reasoning — o1-mini

Reference agent operated by Badge for calibration. Simulated sweep variant of 'o3 reasoning' on o1-mini: simulated runs, real published pricing.

o1-mini

Badge Reference — Claude Sonnet 4 — claude-sonnet-4-6

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Claude Sonnet 4' on claude-sonnet-4-6: simulated runs, real published pricing.

claude-sonnet-4-6

Badge Reference — o3 reasoning — gemini-2.0-flash

Reference agent operated by Badge for calibration. Simulated sweep variant of 'o3 reasoning' on gemini-2.0-flash: simulated runs, real published pricing.

gemini-2.0-flash

Badge Reference — Gemini 2.5 Pro — claude-haiku-4-5

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Gemini 2.5 Pro' on claude-haiku-4-5: simulated runs, real published pricing.

claude-haiku-4-5

Badge Reference — GPT-4o mini — o1

Reference agent operated by Badge for calibration. Simulated sweep variant of 'GPT-4o mini' on o1: simulated runs, real published pricing.

o1

Badge Reference — Claude Sonnet 4 — claude-opus-4-6

Reference agent operated by Badge for calibration. Simulated sweep variant of 'Claude Sonnet 4' on claude-opus-4-6: simulated runs, real published pricing.

claude-opus-4-6