From e6ca55ead8817c8ace49695ce71e2f043cb5eba7 Mon Sep 17 00:00:00 2001 From: thatdevguy <46411187+chrissalomon@users.noreply.github.com> Date: Tue, 13 Oct 2026 11:00:16 -0400 Subject: [PATCH] Add Tempr's Cohere, AI21, Inception, Morph and Sarvam models 17 chat models these five labs' own APIs currently serve, on the caller's own key: Cohere 10, AI21 2, Inception 2, Morph 2, Sarvam 1. The Inception and Morph entries point at the models/ lab files added in #9531. Co-Authored-By: Claude Opus 5.5 --- providers/tempr/models/ai21/jamba-large.toml | 5 +++++ providers/tempr/models/ai21/jamba-mini.toml | 5 +++++ .../tempr/models/cohere/command-a-03-2025.toml | 5 +++++ .../models/cohere/command-a-plus-05-2026.toml | 15 +++++++++++++++ .../cohere/command-a-reasoning-08-2025.toml | 15 +++++++++++++++ .../cohere/command-a-translate-08-2025.toml | 5 +++++ .../models/cohere/command-a-vision-07-2025.toml | 5 +++++ .../tempr/models/cohere/command-r-08-2024.toml | 5 +++++ .../models/cohere/command-r-plus-08-2024.toml | 5 +++++ .../tempr/models/cohere/command-r7b-12-2024.toml | 5 +++++ .../models/cohere/command-r7b-arabic-02-2025.toml | 5 +++++ .../tempr/models/cohere/north-mini-code-1-0.toml | 14 ++++++++++++++ providers/tempr/models/inception/mercury-2.5.toml | 14 ++++++++++++++ providers/tempr/models/inception/mercury-2.toml | 14 ++++++++++++++ providers/tempr/models/morph/morph-v3-fast.toml | 5 +++++ providers/tempr/models/morph/morph-v3-large.toml | 5 +++++ providers/tempr/models/sarvam/sarvam-105b.toml | 11 +++++++++++ 17 files changed, 138 insertions(+) create mode 100644 providers/tempr/models/ai21/jamba-large.toml create mode 100644 providers/tempr/models/ai21/jamba-mini.toml create mode 100644 providers/tempr/models/cohere/command-a-03-2025.toml create mode 100644 providers/tempr/models/cohere/command-a-plus-05-2026.toml create mode 100644 providers/tempr/models/cohere/command-a-reasoning-08-2025.toml create mode 100644 providers/tempr/models/cohere/command-a-translate-08-2025.toml create mode 100644 providers/tempr/models/cohere/command-a-vision-07-2025.toml create mode 100644 providers/tempr/models/cohere/command-r-08-2024.toml create mode 100644 providers/tempr/models/cohere/command-r-plus-08-2024.toml create mode 100644 providers/tempr/models/cohere/command-r7b-12-2024.toml create mode 100644 providers/tempr/models/cohere/command-r7b-arabic-02-2025.toml create mode 100644 providers/tempr/models/cohere/north-mini-code-1-0.toml create mode 100644 providers/tempr/models/inception/mercury-2.5.toml create mode 100644 providers/tempr/models/inception/mercury-2.toml create mode 100644 providers/tempr/models/morph/morph-v3-fast.toml create mode 100644 providers/tempr/models/morph/morph-v3-large.toml create mode 100644 providers/tempr/models/sarvam/sarvam-105b.toml diff --git a/providers/tempr/models/ai21/jamba-large.toml b/providers/tempr/models/ai21/jamba-large.toml new file mode 100644 index 00000000000..e6dbfa4b9f5 --- /dev/null +++ b/providers/tempr/models/ai21/jamba-large.toml @@ -0,0 +1,5 @@ +base_model = "ai21/jamba-large" + +[cost] +input = 2 +output = 8 diff --git a/providers/tempr/models/ai21/jamba-mini.toml b/providers/tempr/models/ai21/jamba-mini.toml new file mode 100644 index 00000000000..4cb2362c163 --- /dev/null +++ b/providers/tempr/models/ai21/jamba-mini.toml @@ -0,0 +1,5 @@ +base_model = "ai21/jamba-mini" + +[cost] +input = 0.2 +output = 0.4 diff --git a/providers/tempr/models/cohere/command-a-03-2025.toml b/providers/tempr/models/cohere/command-a-03-2025.toml new file mode 100644 index 00000000000..35796b07987 --- /dev/null +++ b/providers/tempr/models/cohere/command-a-03-2025.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-a-03-2025" + +[cost] +input = 2.5 +output = 10 diff --git a/providers/tempr/models/cohere/command-a-plus-05-2026.toml b/providers/tempr/models/cohere/command-a-plus-05-2026.toml new file mode 100644 index 00000000000..c795975b22a --- /dev/null +++ b/providers/tempr/models/cohere/command-a-plus-05-2026.toml @@ -0,0 +1,15 @@ +# Tempr Gateway: https://temprhq.io/docs/gateway-chat-completions#reasoning +# Toggle: reasoning.enabled = true|false (reasoning.effort = "none" also turns it off) +# Budget: reasoning.max_tokens (integer reasoning tokens) +# /v1/messages: thinking.type = enabled|disabled; thinking.budget_tokens +# /v1/responses: reasoning.effort = none|low|medium|high (sized to a thinking budget) +base_model = "cohere/command-a-plus-05-2026" + +reasoning_options = [ + { type = "toggle" }, + { type = "budget_tokens", min = 1 }, +] + +[cost] +input = 2.5 +output = 10 diff --git a/providers/tempr/models/cohere/command-a-reasoning-08-2025.toml b/providers/tempr/models/cohere/command-a-reasoning-08-2025.toml new file mode 100644 index 00000000000..09af951e2bb --- /dev/null +++ b/providers/tempr/models/cohere/command-a-reasoning-08-2025.toml @@ -0,0 +1,15 @@ +# Tempr Gateway: https://temprhq.io/docs/gateway-chat-completions#reasoning +# Toggle: reasoning.enabled = true|false (reasoning.effort = "none" also turns it off) +# Budget: reasoning.max_tokens (integer reasoning tokens) +# /v1/messages: thinking.type = enabled|disabled; thinking.budget_tokens +# /v1/responses: reasoning.effort = none|low|medium|high (sized to a thinking budget) +base_model = "cohere/command-a-reasoning-08-2025" + +reasoning_options = [ + { type = "toggle" }, + { type = "budget_tokens", min = 1 }, +] + +[cost] +input = 2.5 +output = 10 diff --git a/providers/tempr/models/cohere/command-a-translate-08-2025.toml b/providers/tempr/models/cohere/command-a-translate-08-2025.toml new file mode 100644 index 00000000000..595f923fbbb --- /dev/null +++ b/providers/tempr/models/cohere/command-a-translate-08-2025.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-a-translate-08-2025" + +[cost] +input = 2.5 +output = 10 diff --git a/providers/tempr/models/cohere/command-a-vision-07-2025.toml b/providers/tempr/models/cohere/command-a-vision-07-2025.toml new file mode 100644 index 00000000000..57c7fc6584d --- /dev/null +++ b/providers/tempr/models/cohere/command-a-vision-07-2025.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-a-vision-07-2025" + +[cost] +input = 2.5 +output = 10 diff --git a/providers/tempr/models/cohere/command-r-08-2024.toml b/providers/tempr/models/cohere/command-r-08-2024.toml new file mode 100644 index 00000000000..5b6e7da0998 --- /dev/null +++ b/providers/tempr/models/cohere/command-r-08-2024.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-r-08-2024" + +[cost] +input = 0.15 +output = 0.6 diff --git a/providers/tempr/models/cohere/command-r-plus-08-2024.toml b/providers/tempr/models/cohere/command-r-plus-08-2024.toml new file mode 100644 index 00000000000..419d99d79cc --- /dev/null +++ b/providers/tempr/models/cohere/command-r-plus-08-2024.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-r-plus-08-2024" + +[cost] +input = 2.5 +output = 10 diff --git a/providers/tempr/models/cohere/command-r7b-12-2024.toml b/providers/tempr/models/cohere/command-r7b-12-2024.toml new file mode 100644 index 00000000000..815678c3b34 --- /dev/null +++ b/providers/tempr/models/cohere/command-r7b-12-2024.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-r7b-12-2024" + +[cost] +input = 0.0375 +output = 0.15 diff --git a/providers/tempr/models/cohere/command-r7b-arabic-02-2025.toml b/providers/tempr/models/cohere/command-r7b-arabic-02-2025.toml new file mode 100644 index 00000000000..460366cb390 --- /dev/null +++ b/providers/tempr/models/cohere/command-r7b-arabic-02-2025.toml @@ -0,0 +1,5 @@ +base_model = "cohere/command-r7b-arabic-02-2025" + +[cost] +input = 0.0375 +output = 0.15 diff --git a/providers/tempr/models/cohere/north-mini-code-1-0.toml b/providers/tempr/models/cohere/north-mini-code-1-0.toml new file mode 100644 index 00000000000..129d9392c3a --- /dev/null +++ b/providers/tempr/models/cohere/north-mini-code-1-0.toml @@ -0,0 +1,14 @@ +# Tempr Gateway: https://temprhq.io/docs/gateway-chat-completions#reasoning +# Off: reasoning.effort = "none" (or reasoning.enabled = false) +# Effort: reasoning.effort = none|high (alias: top-level reasoning_effort) +# /v1/messages: thinking.type = enabled|disabled; output_config.effort = high +# /v1/responses: reasoning.effort = none|high +base_model = "cohere/north-mini-code-1-0" + +reasoning_options = [ + { type = "effort", values = ["none", "high"] }, +] + +[cost] +input = 0 +output = 0 diff --git a/providers/tempr/models/inception/mercury-2.5.toml b/providers/tempr/models/inception/mercury-2.5.toml new file mode 100644 index 00000000000..06b9e0e2766 --- /dev/null +++ b/providers/tempr/models/inception/mercury-2.5.toml @@ -0,0 +1,14 @@ +# Tempr Gateway: https://temprhq.io/docs/gateway-chat-completions#reasoning +# Effort: reasoning.effort = low|medium|high (alias: top-level reasoning_effort) +# /v1/messages: output_config.effort = low|medium|high +# /v1/responses: reasoning.effort = low|medium|high +base_model = "inception/mercury-2.5" + +reasoning_options = [ + { type = "effort", values = ["low", "medium", "high"] }, +] + +[cost] +input = 0.04 +output = 0.15 +cache_read = 0.004 diff --git a/providers/tempr/models/inception/mercury-2.toml b/providers/tempr/models/inception/mercury-2.toml new file mode 100644 index 00000000000..ae9df5856bd --- /dev/null +++ b/providers/tempr/models/inception/mercury-2.toml @@ -0,0 +1,14 @@ +# Tempr Gateway: https://temprhq.io/docs/gateway-chat-completions#reasoning +# Effort: reasoning.effort = low|medium|high (alias: top-level reasoning_effort) +# /v1/messages: output_config.effort = low|medium|high +# /v1/responses: reasoning.effort = low|medium|high +base_model = "inception/mercury-2" + +reasoning_options = [ + { type = "effort", values = ["low", "medium", "high"] }, +] + +[cost] +input = 0.25 +output = 0.75 +cache_read = 0.025 diff --git a/providers/tempr/models/morph/morph-v3-fast.toml b/providers/tempr/models/morph/morph-v3-fast.toml new file mode 100644 index 00000000000..bb8ce4b6cf9 --- /dev/null +++ b/providers/tempr/models/morph/morph-v3-fast.toml @@ -0,0 +1,5 @@ +base_model = "morph/morph-v3-fast" + +[cost] +input = 0.8 +output = 1.2 diff --git a/providers/tempr/models/morph/morph-v3-large.toml b/providers/tempr/models/morph/morph-v3-large.toml new file mode 100644 index 00000000000..cfa7b285e48 --- /dev/null +++ b/providers/tempr/models/morph/morph-v3-large.toml @@ -0,0 +1,5 @@ +base_model = "morph/morph-v3-large" + +[cost] +input = 0.9 +output = 1.9 diff --git a/providers/tempr/models/sarvam/sarvam-105b.toml b/providers/tempr/models/sarvam/sarvam-105b.toml new file mode 100644 index 00000000000..d6f95610113 --- /dev/null +++ b/providers/tempr/models/sarvam/sarvam-105b.toml @@ -0,0 +1,11 @@ +# Tempr Gateway: https://temprhq.io/docs/gateway-chat-completions#reasoning +# No off switch on Tempr: Sarvam switches thinking off with a JSON null effort, which Tempr +# doesn't send, so reasoning.effort = "none" maps to the lowest level below. +# Effort: reasoning.effort = low|medium|high (alias: top-level reasoning_effort) +# /v1/messages: output_config.effort = low|medium|high +# /v1/responses: reasoning.effort = low|medium|high +base_model = "sarvam/sarvam-105b" + +reasoning_options = [ + { type = "effort", values = ["low", "medium", "high"] }, +]