Skip to content
Model infoModelsSnapshot: 15 Oct 2026

NEC / cotomi

cotomi 13B (original)

Original NEC language-model generation announced in July 2023. The historical record keeps its published architecture separate from later cotomi releases.

historicalClosed weightsText generation
Official documentation (opens in a new tab)Reviewed 14 Oct 2026
10 of 22 specification fields collected
Context windowNot collectedAwaiting source review
Maximum outputNot collectedAwaiting source review
Input billingNot collectedNEC (historical generation)
Output billingNot collectedNo inferred price

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
Maximum input tokens
Not collected
Maximum output
Not collected
Reasoning
Not collected
Tool calling
Not collected
Structured output
Not collected
Streaming
Not collected
Batch processing
Not collected
Fine-tuning
Not collected
Knowledge cutoff
Not collected
Release date
2023-07-06Source for Release date (opens in a new tab)

Announcement date; internal use was already underway.

Architecture
Transformer; 40 layers, hidden size 5120, 40 attention headsSource for Architecture (opens in a new tab)

Technical-journal description of the original 13B family. Matching LLaMA dimensions do not establish LLaMA weight lineage.

Languages
Japanese, EnglishSource for Languages (opens in a new tab)

Training includes Japanese, English and source code; no comprehensive locale-quality guarantee.

Context caching
Not collected
Published price
Not collected

No revision-specific public API token tariff verified. Enterprise quotes and appliance licensing are different billing arrangements.

Tokenizer
Byte-pair encoding trained on Japanese and English textSource for Tokenizer (opens in a new tab)
Training sequence length
Initially 2048 tokens; extended through positional interpolationSource for Training sequence length (opens in a new tab)

Training sequence length is not a verified serving context limit.

Alignment method
ChatML instruction tuning and direct preference optimizationSource for Alignment method (opens in a new tab)

Describes the journal’s 13B model; exact deployment checkpoint revision is unpinned.

Hardware requirements
Single-GPU server; cloud or on-premises deploymentSource for Hardware requirements (opens in a new tab)

Access & pricing

Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.

NEC (historical generation)

Provider source (opens in a new tab)
API model ID
Not collected

A public request model string and self-service API contract were not verified. The display name is not an API selector.

Access
Historical model reference. Current availability of the original checkpoint requires provider confirmation.
Regions
Enterprise contract and deployment determine availability and data location.

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Retention, training use, regional processing and enterprise terms have not been summarized for this endpoint.

Caching

Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.

01 / INITIAL INPUTProcess & retain

The system processes eligible input and retains reusable state.

02 / LATER REQUESTCheck available state

Reuse depends on matching input, session state and the cache policy.

03 / CACHE HITReuse & generate

Reusing state can reduce repeated work. Any charges follow the billing rules below.

NEC (historical generation)

Not verified
How it works
Not collected
Minimum reusable input
Not collected
Lifetime & refresh
Not collected
How to enable it
Not collected
Usage fields
Not collected
Billing rules
Not collected
Limits & data boundaries
Not collected

No verified cache rate for this endpoint. An empty rate is not a zero charge.

What the measurements say

No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.

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

NEC describes its proprietary, internally developed model delivered through enterprise products and services. No matching downloadable checkpoint was verified. Closed-weight classification is inferred for this product identity, not for every model developed by NEC.

Classification evidence (opens in a new tab)

Lifecycle dates

Original 13B model announced6 Jul 2023Source (opens in a new tab)

Sources reviewed