Skip to content

Apps token limit #7474

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
DOsinga merged 5 commits into from
Feb 24, 2026
Merged

Apps token limit #7474

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
0815113
Actually use the max output tokens from the different models
Feb 24, 2026
612710c
Update thinking
Feb 24, 2026
1068abd
Remove comment
Feb 24, 2026
2b4ca8b
Update crates/goose/src/providers/formats/anthropic.rs
DOsinga Feb 24, 2026
2f1375b
Update crates/goose/src/model.rs
DOsinga Feb 24, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 2 additions & 4 deletions crates/goose/src/agents/platform_extensions/apps.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -290,8 +290,7 @@ impl AppsManagerClient {
let messages = vec![Message::user().with_text(&user_prompt)];
let tools = vec![Self::create_app_content_tool()];

let mut model_config = provider.get_model_config();
model_config.max_tokens = Some(16384);
let model_config = provider.get_model_config();

let (response, _usage) = provider
.complete(&model_config, session_id, &system_prompt, &messages, &tools)
Expand Down Expand Up @@ -324,8 +323,7 @@ impl AppsManagerClient {
let messages = vec![Message::user().with_text(&user_prompt)];
let tools = vec![Self::update_app_content_tool()];

let mut model_config = provider.get_model_config();
model_config.max_tokens = Some(16384);
let model_config = provider.get_model_config();

let (response, _usage) = provider
.complete(&model_config, session_id, &system_prompt, &messages, &tools)
Expand Down
3 changes: 1 addition & 2 deletions crates/goose/src/model.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -280,8 +280,7 @@ impl ModelConfig {
return tokens;
}

// Priority 2: Global default
4_096
16_384
Comment thread
DOsinga marked this conversation as resolved.
Outdated
}

pub fn new_or_fail(model_name: &str) -> ModelConfig {
Expand Down
10 changes: 6 additions & 4 deletions crates/goose/src/providers/formats/anthropic.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -431,10 +431,12 @@ pub fn create_request(

let is_thinking_enabled = std::env::var("CLAUDE_THINKING_ENABLED").is_ok();
if is_thinking_enabled {
let budget_tokens = std::env::var("CLAUDE_THINKING_BUDGET")
.unwrap_or_else(|_| "16000".to_string())
.parse()
.unwrap_or(16000);
// Anthropic requires budget_tokens >= 1024
const DEFAULT_THINKING_BUDGET: i32 = 16000;
let budget_tokens: i32 = std::env::var("CLAUDE_THINKING_BUDGET")
.ok()
.and_then(|s| s.parse().ok())
.unwrap_or(DEFAULT_THINKING_BUDGET);
Comment thread
DOsinga marked this conversation as resolved.
Outdated

payload
.as_object_mut()
Expand Down
48 changes: 23 additions & 25 deletions crates/goose/src/providers/formats/databricks.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -613,19 +613,19 @@ pub fn create_request(

let is_thinking_enabled = std::env::var("CLAUDE_THINKING_ENABLED").is_ok();
if is_claude_sonnet && is_thinking_enabled {
// Minimum budget_tokens is 1024
let budget_tokens = std::env::var("CLAUDE_THINKING_BUDGET")
.unwrap_or_else(|_| "16000".to_string())
.parse()
.unwrap_or(16000);

// For Claude models with thinking enabled, we need to add max_tokens + budget_tokens
// Default to 8192 (Claude max output) + budget if not specified
let max_completion_tokens = model_config.max_tokens.unwrap_or(8192);
payload.as_object_mut().unwrap().insert(
"max_tokens".to_string(),
json!(max_completion_tokens + budget_tokens),
);
// Anthropic requires budget_tokens >= 1024
const DEFAULT_THINKING_BUDGET: i32 = 16000;
let budget_tokens: i32 = std::env::var("CLAUDE_THINKING_BUDGET")
.ok()
.and_then(|s| s.parse().ok())
.unwrap_or(DEFAULT_THINKING_BUDGET);

Copilot AI Feb 24, 2026

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The comment says Anthropic requires budget_tokens >= 1024, but CLAUDE_THINKING_BUDGET is parsed without enforcing that minimum (or preventing negatives), which can lead to Databricks request rejection; clamp/validate the parsed value to at least 1024 before using it.

Suggested change
.unwrap_or(DEFAULT_THINKING_BUDGET);
.unwrap_or(DEFAULT_THINKING_BUDGET)
.max(1024);

Copilot uses AI. Check for mistakes.

// With thinking enabled, max_tokens must include both output and thinking budget
let max_tokens = model_config.max_output_tokens() + budget_tokens;
payload
.as_object_mut()
.unwrap()
.insert("max_tokens".to_string(), json!(max_tokens));
Comment on lines +623 to +628

Copilot AI Feb 24, 2026

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

model_config.max_output_tokens() + budget_tokens is an i32 addition that can overflow if CLAUDE_THINKING_BUDGET is set large, which can wrap to a negative max_tokens in release builds; use checked/saturating arithmetic (or compute in i64 and clamp) before serializing.

Copilot uses AI. Check for mistakes.

@DOsinga DOsinga Feb 24, 2026

Copy link
Copy Markdown
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

i32 is 2 billion


payload.as_object_mut().unwrap().insert(
"thinking".to_string(),
Expand All @@ -650,18 +650,16 @@ pub fn create_request(
}
}

// open ai reasoning models use max_completion_tokens instead of max_tokens
if let Some(tokens) = model_config.max_tokens {
let key = if is_openai_reasoning_model {
"max_completion_tokens"
} else {
"max_tokens"
};
payload
.as_object_mut()
.unwrap()
.insert(key.to_string(), json!(tokens));
}
// OpenAI reasoning models use max_completion_tokens instead of max_tokens
let key = if is_openai_reasoning_model {
"max_completion_tokens"
} else {
"max_tokens"
};
payload
.as_object_mut()
.unwrap()
.insert(key.to_string(), json!(model_config.max_output_tokens()));
}

// Apply cache control for Claude models to enable prompt caching
Expand Down
17 changes: 5 additions & 12 deletions crates/goose/src/providers/formats/google.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -592,18 +592,11 @@ pub fn create_request(

let thinking_config = get_thinking_config(model_config);

let generation_config = if model_config.temperature.is_some()
|| model_config.max_tokens.is_some()
|| thinking_config.is_some()
{
Some(GenerationConfig {
temperature: model_config.temperature.map(|t| t as f64),
max_output_tokens: model_config.max_tokens,
thinking_config,
})
} else {
None
};
let generation_config = Some(GenerationConfig {
temperature: model_config.temperature.map(|t| t as f64),
max_output_tokens: Some(model_config.max_output_tokens()),
thinking_config,
});

let request = GoogleRequest {
system_instruction: SystemInstruction {
Expand Down
22 changes: 10 additions & 12 deletions crates/goose/src/providers/formats/openai.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -828,18 +828,16 @@ pub fn create_request(
}
}

// o1 models use max_completion_tokens instead of max_tokens
if let Some(tokens) = model_config.max_tokens {
let key = if is_ox_model {
"max_completion_tokens"
} else {
"max_tokens"
};
payload
.as_object_mut()
.unwrap()
.insert(key.to_string(), json!(tokens));
}
// o1/o3 models use max_completion_tokens instead of max_tokens
let key = if is_ox_model {
"max_completion_tokens"
} else {
"max_tokens"
};
payload
.as_object_mut()
.unwrap()
.insert(key.to_string(), json!(model_config.max_output_tokens()));

if for_streaming {
payload["stream"] = json!(true);
Expand Down
10 changes: 4 additions & 6 deletions crates/goose/src/providers/formats/openai_responses.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -460,12 +460,10 @@ pub fn create_responses_request(
.insert("temperature".to_string(), json!(temp));
}

if let Some(tokens) = model_config.max_tokens {
payload
.as_object_mut()
.unwrap()
.insert("max_output_tokens".to_string(), json!(tokens));
}
payload.as_object_mut().unwrap().insert(
"max_output_tokens".to_string(),
json!(model_config.max_output_tokens()),
);

Ok(payload)
}
Expand Down
3 changes: 1 addition & 2 deletions crates/goose/src/providers/formats/snowflake.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -353,11 +353,10 @@ pub fn create_request(
format_tools(tools)
};

let max_tokens = model_config.max_tokens.unwrap_or(4096);
let mut payload = json!({
"model": model_config.model_name,
"messages": snowflake_messages,
"max_tokens": max_tokens,
"max_tokens": model_config.max_output_tokens(),
});

// Add tools if present and not a description request
Expand Down
2 changes: 1 addition & 1 deletion crates/goose/src/providers/sagemaker_tgi.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -149,7 +149,7 @@ impl SageMakerTgiProvider {
let request = json!({
"inputs": prompt,
"parameters": {
"max_new_tokens": self.model.max_tokens.unwrap_or(150),
"max_new_tokens": self.model.max_output_tokens(),
"temperature": self.model.temperature.unwrap_or(0.7),
Comment on lines 151 to 153

Copilot AI Feb 24, 2026

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Switching TGI's max_new_tokens from a small provider-specific default (150) to model.max_output_tokens() means an unset/unknown model will now request 16,384 tokens by default, which is likely to cause slow/expensive generations or SageMaker endpoint errors; consider keeping a conservative TGI fallback (e.g., use max_tokens only when explicitly set, otherwise default to ~150 or a configurable env var).

Copilot uses AI. Check for mistakes.
"do_sample": true,
"return_full_text": false
Expand Down
Loading