diff --git a/config.yaml b/config.yaml index c2abf7a..c4a13ee 100644 --- a/config.yaml +++ b/config.yaml @@ -6,6 +6,8 @@ # via environment variable TINFOIL_API_KEY. # # Pricing (model_info) is per-token, derived from provider per-1M rates: +# EUR->USD FX = 1.16 (policy, applied 2026-11-18; GreenPT/Regolo quote EUR). +# LiteLLM's cost fields are USD; EUR list prices are converted here, never billed raw. # greenpt/deepseek-v4.1-flash: $0.36 in / $1.45 out / $0.04 cached per 1M (1017 snapshot) # tinfoil/deepseek-v4-1-flash: $0.65 in / $1.45 out / $0.13 cached per 1M # tinfoil/gpt-oss-120b: $0.15 in / $0.60 out per 1M @@ -42,9 +44,9 @@ model_list: api_base: https://api.greenpt.ai/v1 api_key: os.environ/GREENPT_API_KEY model_info: - input_cost_per_token: 0.00000036 - output_cost_per_token: 0.00000145 - cache_read_input_token_cost: 0.00000004 + input_cost_per_token: 3.6e-07 + output_cost_per_token: 1.45e-06 + cache_read_input_token_cost: 4e-08 # DeepSeek V4.1 Flash on Tinfoil (TEE) — private counterpart to the default. - model_name: tinfoil/deepseek-v4-1-flash @@ -93,9 +95,9 @@ model_list: api_base: https://api.greenpt.ai/v1 api_key: os.environ/GREENPT_API_KEY model_info: - input_cost_per_token: 0.00000011 - output_cost_per_token: 0.00000044 - cache_read_input_token_cost: 0.000000022 + input_cost_per_token: 1.1e-07 + output_cost_per_token: 4.4e-07 + cache_read_input_token_cost: 2.2e-08 # GLM-5.3 (full flagship) on GreenPT — strongest agentic model in the green # tier. $1.10 in / $4.40 out per 1M, 1M context. @@ -105,9 +107,9 @@ model_list: api_base: https://api.greenpt.ai/v1 api_key: os.environ/GREENPT_API_KEY model_info: - input_cost_per_token: 0.0000011 - output_cost_per_token: 0.0000044 - cache_read_input_token_cost: 0.000000275 + input_cost_per_token: 1.1e-06 + output_cost_per_token: 4.4e-06 + cache_read_input_token_cost: 2.75e-07 # --- Audio models (Tinfoil, inside the TEE) --- # NOT chat models: these power /v1/audio/* routes (STT for dictation, TTS for @@ -162,8 +164,8 @@ model_list: api_key: os.environ/REGOLO_API_KEY model_info: mode: chat - input_cost_per_token: 2.0e-7 - output_cost_per_token: 7.0e-7 + input_cost_per_token: 2e-07 + output_cost_per_token: 7e-07 # Gemma 4 (31B) — Google's dense multimodal Gemma 4. $0.40 in / $0.60 out # per 1M, 100K ctx. @@ -174,8 +176,8 @@ model_list: api_key: os.environ/REGOLO_API_KEY model_info: mode: chat - input_cost_per_token: 4.0e-7 - output_cost_per_token: 6.0e-7 + input_cost_per_token: 4e-07 + output_cost_per_token: 6e-07 - model_name: regolo/qwen-image litellm_params: @@ -305,8 +307,8 @@ model_list: api_key: os.environ/GREENPT_API_KEY model_info: mode: embedding - input_cost_per_token: 0.00000025 - output_cost_per_token: 0.0 + input_cost_per_token: 2.5e-07 + output_cost_per_token: 0 max_input_tokens: 32000 # --- Embedding model (GreenPT, member-facing) --- @@ -326,8 +328,8 @@ model_list: api_key: os.environ/GREENPT_API_KEY model_info: mode: embedding - input_cost_per_token: 0.0000002 - output_cost_per_token: 0.0 + input_cost_per_token: 2e-07 + output_cost_per_token: 0 max_input_tokens: 32000 # Fallback: if DeepSeek is down, use GPT-OSS