Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Snowflake

Enterprise data platform with proprietary Arctic-Extract document models and managed Cortex services. Its public-weight research releases are excluded from closed-weight coverage.

Arctic-Extract, inferred-private SQL models and explicitly managed task-model aliases reviewed. Public weights are excluded. Research-to-production mappings, exact versions and contractual details remain incomplete.

Closed-weight models
Browse by model typeCapabilities can overlap

Start with the job your model needs to do. Then narrow by capabilities, provider and price.

Select up to 3 models to compare
What does one question cost?

Compare every model with the same request.

Estimate = (input tokens × input price + output tokens × output price) / 1,000,000. Include conversation history in input and billed reasoning in output. Input-only decision models ignore the output setting; enter all billed input for one decision, including its question and criteria. Media, tools and discounts are separate.

10 models

Model input and output limits, prices and estimated cost per question at the selected token counts.
CompareModelInput tokensPublished limit*Output tokensMaximum per response*Input pricePer 1M tokensOutput pricePer 1M tokensCost per question1,000 input + 500 outputEstimated · uncached
125 pages at scale 1; 62 at scale 2; 50 at scale 2.5; 31 at scale 4document page limits512 tokens per entity question; 4,096 tokens per table answeroutput limits by taskUsage pricing
5.55 credits / 1M processed tokensBase extraction · AI Credits · Snowflake AI_EXTRACT9.14 credits / 1M processed tokensFine-tuned extraction · AI Credits · Snowflake AI_EXTRACT fine-tunedPage tokens + prompt tokens + output tokens. Document page baseline: 970 tokens, multiplied by scale_factor.Rates & conditions ↗
Different billing unitPrice details ↗
Not collectedEntity and table extraction JSON; confidence scores are not supported.output formatsUsage pricing
9.5 credits / 1M processed tokensLegacy entity extraction · AI Credits · Snowflake legacy AI_EXTRACT28.4 credits / 1M processed tokensLegacy table extraction · AI Credits · Snowflake legacy AI_EXTRACTRates & conditions ↗
Task-based pricingPrice details ↗
734,003,200input content limitJSON transcript; optional word segments or speaker turns with start/end times.output formatsUsage pricing
0.97 credits / 1M audio tokensTranscription · AI Credits · Snowflake Cortex SQL50 billable tokens per second; minimum 10 seconds per file. One hour = 180,000 tokens.Rates & conditions ↗
Task-based pricingPrice details ↗
100,000max input tokens100,000max output tokensUsage pricing
1.63 credits / 1M processed tokensTranslation · AI Credits · Snowflake Cortex SQLRates & conditions ↗
Task-based pricingPrice details ↗
2,048max input tokensOverall and requested aspect labels: positive, negative, neutral, mixed or unknown.output formatsUsage pricing
1.6 credits / 1M processed tokensSentiment · AI Credits · Snowflake Cortex SQLRates & conditions ↗
Task-based pricingPrice details ↗
Up to 2,000 pages per documentdocument page limitsOCR text or layout-preserving Markdown in JSON; optional per-page results and extracted images.output formatsUsage pricing
0.68 credits / 1,000 pagesOCR parsing · AI Credits · Snowflake Cortex SQL3.66 credits / 1,000 pagesLAYOUT parsing · AI Credits · Snowflake Cortex SQLPDF/DOCX per page; JPEG/JPG/TIF/TIFF/PNG per image file; HTML/TXT per started 3,000-character chunk.Rates & conditions ↗
Task-based pricingPrice details ↗
2,048max input tokensAnswer string extracted from the supplied plain-English or JSON text.output formatsUsage pricingNot collectedTask-based pricingPrice details ↗
Not collectedSnowflake SQLoutput formatsUsage pricingNot collectedNeeds rate detailsPrice details ↗
Not collectedInline Snowflake SQL completionsoutput formatsUsage pricingNot collectedNeeds rate detailsPrice details ↗
Not collectedInline Snowflake SQL completionsoutput formatsUsage pricingNot collectedNeeds rate detailsPrice details ↗

* Media models show their native limits. A shared context window includes both input and output tokens.

Original currency and billing unit, without currency conversion. Rates are dated snapshots; open the conditions for tiers, promotions and endpoint restrictions.How to read the data

Catalog coverage

Closed-weight model variants and explicitly identified closed-source managed model components across language, code, image, video, audio and other documented modalities. Global provider and historical-release coverage is being expanded; this is not yet exhaustive. 172 company reviews are still open.

See what is being reviewed