Input / 1M tokens$3.00Devin · SWE-2 High · Self-serve
Output / 1M tokens$15.00USD · observed 15 Oct 2026
Specifications
Provider-documented facts. “Not collected” means we have not verified that field yet, not that the feature is unsupported.
Context / input limit
Not collected
An exact-version public contract was not established. Agent capabilities, underlying base-model limits and provider-side RL are not standalone model API guarantees.
Maximum input tokens
Not collected
An exact-version public contract was not established. Agent capabilities, underlying base-model limits and provider-side RL are not standalone model API guarantees.
Maximum output
Not collected
An exact-version public contract was not established. Agent capabilities, underlying base-model limits and provider-side RL are not standalone model API guarantees.
Parameters
Not collected
An exact-version public contract was not established. Agent capabilities, underlying base-model limits and provider-side RL are not standalone model API guarantees.
Used within the provider agent harness; this does not establish arbitrary public function-schema support.
Structured output
Not collected
An exact-version public contract was not established. Agent capabilities, underlying base-model limits and provider-side RL are not standalone model API guarantees.
Streaming
Not collected
An exact-version public contract was not established. Agent capabilities, underlying base-model limits and provider-side RL are not standalone model API guarantees.
Batch processing
Not collected
An exact-version public contract was not established. Agent capabilities, underlying base-model limits and provider-side RL are not standalone model API guarantees.
Fine-tuning
Not collected
An exact-version public contract was not established. Agent capabilities, underlying base-model limits and provider-side RL are not standalone model API guarantees.
Distinct cached-input rates are published for the model; eligibility and rates vary by tier.
Knowledge cutoff
Not collected
An exact-version public contract was not established. Agent capabilities, underlying base-model limits and provider-side RL are not standalone model API guarantees.
Published Self-serve SWE-2 High model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Scheduled list rates after the published promotion. Cached input price: $0.30 / 1M tokens (base tier).
Self-serve data may be used for model training by default. Paid customers can opt out; Teams requires an admin. Enterprise data is not used for training without prior written consent. Model-provider zero retention after opt-out is not zero retention of all Devin account/session data; customer data and feedback have separate retention rules.
Published model UID in the product catalogue. This is not proof of a public raw inference endpoint.
Access
Desktop/CLI product model selection. Account entitlements and team restrictions apply. Published dollar token rates are converted to contract-specific ACUs.
Regions
Model-specific processing region not established; dedicated enterprise deployment options depend on the agreement.
Published Enterprise ACUs SWE-2 High model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Promotion valid through 2026-12-31 inclusive; publisher timezone not specified. Cached input price: $0.075 / 1M tokens (base tier).
Self-serve data may be used for model training by default. Paid customers can opt out; Teams requires an admin. Enterprise data is not used for training without prior written consent. Model-provider zero retention after opt-out is not zero retention of all Devin account/session data; customer data and feedback have separate retention rules.
Published model UID in the product catalogue. This is not proof of a public raw inference endpoint.
Access
Only legacy credit-based enterprise plans. Prompt credits are charged per message to a premium model, not per million tokens.
Regions
Model-specific processing region not established; dedicated enterprise deployment options depend on the agreement.
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 15 Oct 2026
Category
Usage
Rate
Unit
Legacy enterprise message pricing
Prompt message1
9 credits
message
Rate conditions
Only legacy enterprise credit billing. Different effort and speed selections have different message multipliers. These credits are not ACUs, dollars or token-based cost-per-question quotes.
Enterprise documentation states no model training on customer data by default. Dedicated deployments store customer data in the customer tenant; retention and contractual deletion are distinct from model-provider cache policies.
Published Self-serve SWE-2 Medium model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Scheduled list rates after the published promotion. Cached input price: $0.30 / 1M tokens (base tier).
Self-serve data may be used for model training by default. Paid customers can opt out; Teams requires an admin. Enterprise data is not used for training without prior written consent. Model-provider zero retention after opt-out is not zero retention of all Devin account/session data; customer data and feedback have separate retention rules.
Published model UID in the product catalogue. This is not proof of a public raw inference endpoint.
Access
Desktop/CLI product model selection. Account entitlements and team restrictions apply. Published dollar token rates are converted to contract-specific ACUs.
Regions
Model-specific processing region not established; dedicated enterprise deployment options depend on the agreement.
Published Enterprise ACUs SWE-2 Medium model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Promotion valid through 2026-12-31 inclusive; publisher timezone not specified. Cached input price: $0.075 / 1M tokens (base tier).
Self-serve data may be used for model training by default. Paid customers can opt out; Teams requires an admin. Enterprise data is not used for training without prior written consent. Model-provider zero retention after opt-out is not zero retention of all Devin account/session data; customer data and feedback have separate retention rules.
Published model UID in the product catalogue. This is not proof of a public raw inference endpoint.
Access
Only legacy credit-based enterprise plans. Prompt credits are charged per message to a premium model, not per million tokens.
Regions
Model-specific processing region not established; dedicated enterprise deployment options depend on the agreement.
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 15 Oct 2026
Category
Usage
Rate
Unit
Legacy enterprise message pricing
Prompt message1
6 credits
message
Rate conditions
Only legacy enterprise credit billing. Different effort and speed selections have different message multipliers. These credits are not ACUs, dollars or token-based cost-per-question quotes.
Enterprise documentation states no model training on customer data by default. Dedicated deployments store customer data in the customer tenant; retention and contractual deletion are distinct from model-provider cache policies.
Published Self-serve SWE-2 Max model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Scheduled list rates after the published promotion. Cached input price: $0.30 / 1M tokens (base tier).
Self-serve data may be used for model training by default. Paid customers can opt out; Teams requires an admin. Enterprise data is not used for training without prior written consent. Model-provider zero retention after opt-out is not zero retention of all Devin account/session data; customer data and feedback have separate retention rules.
Published model UID in the product catalogue. This is not proof of a public raw inference endpoint.
Access
Desktop/CLI product model selection. Account entitlements and team restrictions apply. Published dollar token rates are converted to contract-specific ACUs.
Regions
Model-specific processing region not established; dedicated enterprise deployment options depend on the agreement.
Published Enterprise ACUs SWE-2 Max model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Promotion valid through 2026-12-31 inclusive; publisher timezone not specified. Cached input price: $0.075 / 1M tokens (base tier).
Self-serve data may be used for model training by default. Paid customers can opt out; Teams requires an admin. Enterprise data is not used for training without prior written consent. Model-provider zero retention after opt-out is not zero retention of all Devin account/session data; customer data and feedback have separate retention rules.
Published model UID in the product catalogue. This is not proof of a public raw inference endpoint.
Access
Only legacy credit-based enterprise plans. Prompt credits are charged per message to a premium model, not per million tokens.
Regions
Model-specific processing region not established; dedicated enterprise deployment options depend on the agreement.
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 15 Oct 2026
Category
Usage
Rate
Unit
Legacy enterprise message pricing
Prompt message1
12 credits
message
Rate conditions
Only legacy enterprise credit billing. Different effort and speed selections have different message multipliers. These credits are not ACUs, dollars or token-based cost-per-question quotes.
Enterprise documentation states no model training on customer data by default. Dedicated deployments store customer data in the customer tenant; retention and contractual deletion are distinct from model-provider cache policies.
Exact cache creation/matching controls not documented here.
Usage fields
Account model-discovery responses include cached_input/cache_read/cache_write price dimensions. These are prices, not response usage counters.Source for Usage fields (opens in a new tab)
Published Self-serve SWE-2 High model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Scheduled list rates after the published promotion.
Observed 15 Oct 2026. Write rates and storage fees are different charges; use the billing rules above.
Exact cache creation/matching controls not documented here.
Usage fields
Account model-discovery responses include cached_input/cache_read/cache_write price dimensions. These are prices, not response usage counters.Source for Usage fields (opens in a new tab)
Published Enterprise ACUs SWE-2 High model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Promotion valid through 2026-12-31 inclusive; publisher timezone not specified.
Observed 15 Oct 2026. Write rates and storage fees are different charges; use the billing rules above.
Caching reviewed 15 Oct 2026
Devin · SWE-2 High · Legacy credits
swe-2-high
Not verified
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
How it works
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Minimum reusable input
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Lifetime & refresh
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
How to enable it
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Usage fields
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Billing rules
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Limits & data boundaries
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Exact cache creation/matching controls not documented here.
Usage fields
Account model-discovery responses include cached_input/cache_read/cache_write price dimensions. These are prices, not response usage counters.Source for Usage fields (opens in a new tab)
Published Self-serve SWE-2 Medium model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Scheduled list rates after the published promotion.
Observed 15 Oct 2026. Write rates and storage fees are different charges; use the billing rules above.
Exact cache creation/matching controls not documented here.
Usage fields
Account model-discovery responses include cached_input/cache_read/cache_write price dimensions. These are prices, not response usage counters.Source for Usage fields (opens in a new tab)
Published Enterprise ACUs SWE-2 Medium model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Promotion valid through 2026-12-31 inclusive; publisher timezone not specified.
Observed 15 Oct 2026. Write rates and storage fees are different charges; use the billing rules above.
Caching reviewed 15 Oct 2026
Devin · SWE-2 Medium · Legacy credits
swe-2-medium
Not verified
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
How it works
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Minimum reusable input
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Lifetime & refresh
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
How to enable it
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Usage fields
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Billing rules
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Limits & data boundaries
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Exact cache creation/matching controls not documented here.
Usage fields
Account model-discovery responses include cached_input/cache_read/cache_write price dimensions. These are prices, not response usage counters.Source for Usage fields (opens in a new tab)
Published Self-serve SWE-2 Max model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Scheduled list rates after the published promotion.
Observed 15 Oct 2026. Write rates and storage fees are different charges; use the billing rules above.
Exact cache creation/matching controls not documented here.
Usage fields
Account model-discovery responses include cached_input/cache_read/cache_write price dimensions. These are prices, not response usage counters.Source for Usage fields (opens in a new tab)
Published Enterprise ACUs SWE-2 Max model-token subtotal; plan/contract and tool or cloud-agent infrastructure charges are separate. Account List Models API is authoritative for actual entitlement and price. Promotion valid through 2026-12-31 inclusive; publisher timezone not specified.
Observed 15 Oct 2026. Write rates and storage fees are different charges; use the billing rules above.
Caching reviewed 15 Oct 2026
Devin · SWE-2 Max · Legacy credits
swe-2-max
Not verified
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
How it works
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Minimum reusable input
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Lifetime & refresh
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
How to enable it
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Usage fields
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Billing rules
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
Limits & data boundaries
Not collected
Exact-version cache support and billing contract not established. Internal KV reuse and retrieved code are not proof of discounted prompt caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 15 Oct 2026
What the measurements say
FrontierCode 1.1 Main
Cognition · provider-reported
50%
September 10, 2026 release table; provider evaluation and harness. Not an independent cross-provider ranking or a guarantee for every reasoning effort.
September 10, 2026 release table; provider evaluation and harness. Not an independent cross-provider ranking or a guarantee for every reasoning effort.
September 10, 2026 release table; provider evaluation and harness. Not an independent cross-provider ranking or a guarantee for every reasoning effort.
September 10, 2026 release table; provider evaluation and harness. Not an independent cross-provider ranking or a guarantee for every reasoning effort.
Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.
The record behind the model
Why it is included
Provider describes a distinct in-house trained model supplied through Windsurf/Devin. Private-weight classification is inferred from the product/research distribution and proprietary service terms; the reviewed sources do not publish an exact checkpoint weight license. A public pretrained base is not a release of the SWE derivative. This inference must be revisited if derivative weights are released.
