Kujtim Hoxha created
cmd/openrouter/main.go | 2 +-
internal/providers/configs/azure.json | 28 +++++++++++++++++++---------
internal/providers/configs/openai.json | 22 ++++++++++++++++------
pkg/provider/provider.go | 22 ++++++++++++----------
4 files changed, 48 insertions(+), 26 deletions(-)
@@ -167,7 +167,7 @@ func main() {
SupportsImages: supportsImages,
}
if model.TopProvider.MaxCompletionTokens != nil {
- m.DefaultMaxTokens = int64(*model.TopProvider.MaxCompletionTokens / 2)
+ m.DefaultMaxTokens = *model.TopProvider.MaxCompletionTokens / 2
} else {
m.DefaultMaxTokens = model.ContextLength / 10
}
@@ -16,7 +16,9 @@
"cost_per_1m_out_cached": 0.375,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": true
},
{
@@ -28,7 +30,9 @@
"cost_per_1m_out_cached": 0.275,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": true
},
{
@@ -40,7 +44,9 @@
"cost_per_1m_out_cached": 0.5,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": true
},
{
@@ -52,7 +58,9 @@
"cost_per_1m_out_cached": 0,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": true
},
{
@@ -63,7 +71,7 @@
"cost_per_1m_in_cached": 0,
"cost_per_1m_out_cached": 0.5,
"context_window": 1047576,
- "default_max_tokens": 20000,
+ "default_max_tokens": 50000,
"can_reason": false,
"supports_attachments": true
},
@@ -75,7 +83,7 @@
"cost_per_1m_in_cached": 0,
"cost_per_1m_out_cached": 0.09999999999999999,
"context_window": 1047576,
- "default_max_tokens": 20000,
+ "default_max_tokens": 50000,
"can_reason": false,
"supports_attachments": true
},
@@ -87,7 +95,7 @@
"cost_per_1m_in_cached": 0,
"cost_per_1m_out_cached": 0.024999999999999998,
"context_window": 1047576,
- "default_max_tokens": 20000,
+ "default_max_tokens": 50000,
"can_reason": false,
"supports_attachments": true
},
@@ -99,7 +107,7 @@
"cost_per_1m_in_cached": 0,
"cost_per_1m_out_cached": 37.5,
"context_window": 128000,
- "default_max_tokens": 20000,
+ "default_max_tokens": 50000,
"can_reason": false,
"supports_attachments": true
},
@@ -112,7 +120,9 @@
"cost_per_1m_out_cached": 0.55,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": false
},
{
@@ -4,7 +4,7 @@
"type": "openai",
"api_key": "$OPENAI_API_KEY",
"api_endpoint": "$OPENAI_API_ENDPOINT",
- "default_large_model_id": "o4-mini",
+ "default_large_model_id": "codex-mini-latest",
"default_small_model_id": "gpt-4o",
"models": [
{
@@ -16,7 +16,9 @@
"cost_per_1m_out_cached": 0.375,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": true
},
{
@@ -28,7 +30,9 @@
"cost_per_1m_out_cached": 0.275,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": true
},
{
@@ -40,7 +44,9 @@
"cost_per_1m_out_cached": 0.5,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": true
},
{
@@ -52,7 +58,9 @@
"cost_per_1m_out_cached": 0,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": true
},
{
@@ -112,7 +120,9 @@
"cost_per_1m_out_cached": 0.55,
"context_window": 200000,
"default_max_tokens": 50000,
- "can_reason": false,
+ "can_reason": true,
+ "has_reasoning_effort": true,
+ "default_reasoning_effort": "medium",
"supports_attachments": false
},
{
@@ -45,16 +45,18 @@ type Provider struct {
// Model represents an AI model configuration.
type Model struct {
- ID string `json:"id"`
- Name string `json:"model"`
- CostPer1MIn float64 `json:"cost_per_1m_in"`
- CostPer1MOut float64 `json:"cost_per_1m_out"`
- CostPer1MInCached float64 `json:"cost_per_1m_in_cached"`
- CostPer1MOutCached float64 `json:"cost_per_1m_out_cached"`
- ContextWindow int64 `json:"context_window"`
- DefaultMaxTokens int64 `json:"default_max_tokens"`
- CanReason bool `json:"can_reason"`
- SupportsImages bool `json:"supports_attachments"`
+ ID string `json:"id"`
+ Name string `json:"model"`
+ CostPer1MIn float64 `json:"cost_per_1m_in"`
+ CostPer1MOut float64 `json:"cost_per_1m_out"`
+ CostPer1MInCached float64 `json:"cost_per_1m_in_cached"`
+ CostPer1MOutCached float64 `json:"cost_per_1m_out_cached"`
+ ContextWindow int64 `json:"context_window"`
+ DefaultMaxTokens int64 `json:"default_max_tokens"`
+ CanReason bool `json:"can_reason"`
+ HasReasoningEffort bool `json:"has_reasoning_efforts"`
+ DefaultReasoningEffort string `json:"default_reasoning_effort,omitempty"`
+ SupportsImages bool `json:"supports_attachments"`
}
// KnownProviders returns all the known inference providers.