SnowPro Specialty: Gen AI (GES-C01)
| Cert name | SnowPro Specialty: Gen AI |
| Exam code | GES-C01 |
| Issued by | Snowflake |
| Service line | AI Service |
| Cost | $375 USD per attempt |
| Duration | 115 minutes |
| Questions | 60–65 (multiple-choice, multiple-select) |
| Passing score | 750 / 1000 |
| Prerequisites | Active SnowPro Core certification (COF-C03) — required |
| Back to cert index | README.md |
Updated exam notice (March 1, 2026): An updated version of the GES-C01 exam was released on March 1, 2026. The Document AI domain has been removed. Anyone sitting the exam on or after March 1 takes the updated version. The guidance on this page reflects the updated exam — do not spend significant time on Document AI.
Why this certification
Snowflake Cortex is increasingly how AI Service delivers Gen AI solutions to clients — LLM inference, semantic search, and RAG pipelines built on a governed, client-controlled data environment. The SnowPro Specialty: Gen AI cert is the credential that proves the AI Service team can execute this work at a production level.
It also signals Snowflake partnership depth: having certified Gen AI practitioners on the team strengthens Brainforge’s positioning in deals where clients are on Snowflake and considering AI feature builds.
Prerequisites
- SnowPro Core must be active — you cannot register without it. Complete snowpro-core.md first and allow a few business days for the certification to process.
- 1+ years of Gen AI experience with Snowflake in an enterprise environment (Snowflake recommendation)
- SQL proficiency; Python experience beneficial
- Hands-on experience with Snowflake Cortex functions (
COMPLETE,SUMMARIZE,TRANSLATE, etc.)
Step 1: Download the study guide
Exam page and registration: learn.snowflake.com/en/certifications/snowpro-GenAI
Updated exam FAQs (March 2026): GenAI-GES-C01-FAQs.pdf
Review the exam FAQs to confirm the current domain list and scope under the updated exam. The Document AI section is no longer tested.
Step 2: Book your exam date
Register at: cp.certmetrics.com/snowflake/en/login
You can only register after your SnowPro Core certificate is active.
Payment note (TBD): The certification exam is paid and typically requires upfront payment at scheduling/checkout time. Brainforge’s policy on who should pay upfront (and how reimbursement/coverage works) is TBD.
Step 3: Complete the prep courses
Snowflake University (free, self-paced)
| # | Course | Where | Notes |
|---|---|---|---|
| 1 | Snowflake x Gen AI: A Cool Collab | learn.snowflake.com/en/courses/OD-SXGA/ | Practical intro to calling an LLM API from Snowflake and building a simple Streamlit UI |
| 2 | Snowflake X GenAI: LLM Functions | learn.snowflake.com/en/courses/OD-SXGA2/ | Hands-on with LLM functions like TRANSLATE, SENTIMENT, SUMMARIZE |
| 3 | Data Science Workshop (Badge 6) | learn.snowflake.com/en/courses/OD-ESS-DSCW/ | Cortex Playground + Cortex Analyst + prompt engineering + applied AI/ML in Snowflake |
External resources
| Resource | Link | Notes |
|---|---|---|
| Medium tutorial: SnowPro Gen AI Part 1 | medium.com/snowflake — Part 1 | Exam info and starting with Snowflake Cortex — read this first for exam orientation |
| DataEngineerHub: How I Passed SnowPro Gen AI 2026 | dataengineerhub.blog | Community-written passing strategy for the current exam version |
| Snowflake Cortex documentation | docs.snowflake.com/en/user-guide/snowflake-cortex/llm-functions | Official function reference — read before the exam, not just during study |
| Snowflake blog: Cortex + Omni | exploreomni.com/blog/snowflake-cortex-and-omni | Contextual reading on how Cortex and Omni work together — relevant to Brainforge’s client stack |
Step 4: Study the exam domains
| Domain | What to study |
|---|---|
| Snowflake Gen AI & LLM Functions | Cortex SQL functions: COMPLETE, SUMMARIZE, TRANSLATE, EXTRACT_ANSWER, SENTIMENT, CLASSIFY_TEXT; vector functions: VECTOR_COSINE_SIMILARITY, VECTOR_INNER_PRODUCT; embedding generation; semantic search patterns; prompt construction in SQL |
| Snowflake Gen AI Overview | Cortex AI architecture, supported LLMs (Mistral, Llama, Snowflake Arctic), Snowflake Copilot, Model Registry, Snowpark Container Services for custom models |
| Snowflake Gen AI Governance | RBAC for Cortex feature usage, CORTEX_FUNCTIONS_USAGE_HISTORY view, cost attribution and budgeting, data residency and privacy considerations for LLM calls |
| Document AI (removed) | Removed from exam as of March 1, 2026 — skip this domain entirely |
What to go deep on for AI Service
The Gen AI & LLM Functions domain is where the exam is won or lost. Every AI Service team member should be able to write a working CORTEX.COMPLETE() call, understand when to use SUMMARIZE vs EXTRACT_ANSWER, and explain how vector similarity search enables a RAG retrieval step — in SQL, not just conceptually.
The Governance domain is easy to underestimate. Clients will ask about cost controls and data privacy for LLM calls. Know CORTEX_FUNCTIONS_USAGE_HISTORY and RBAC for Cortex access by role.
Step 5: Hands-on practice before the exam
Do not sit this exam without having run Cortex functions in a live Snowflake environment. Before exam day:
- Run at least five different Cortex functions (
COMPLETE,SUMMARIZE,TRANSLATE,SENTIMENT,EXTRACT_ANSWER) in a Snowflake trial or sandbox account - Build a simple semantic search example: generate embeddings for a small text dataset, store as a VECTOR column, query with
VECTOR_COSINE_SIMILARITY - Query
CORTEX_FUNCTIONS_USAGE_HISTORYand interpret the results - Articulate in plain English what a RAG pipeline in Snowflake looks like end-to-end (retrieve → embed → complete)
Step 6: Sit the exam
Exam day checklist:
- SnowPro Core cert is active in your Certmetrics profile — verify the day before
- Government-issued photo ID ready
- Quiet, uninterrupted space (online proctored via OnVUE)
- System check completed 24 hours in advance
- 115 minutes for 60–65 questions — roughly 1.5–2 minutes per question; use the extra time on scenario questions
- Skip Document AI questions if any appear — flag and move on; they should not appear on the updated exam
Exam portal: cp.certmetrics.com/snowflake/en/login
After passing
- Share your result in
#ai-learning— log against M1.2 (AI Service milestone). - Your digital badge will arrive via email within a few business days.
- SnowPro Specialty: Gen AI + Claude Certified Architect: Foundations together represent the core Snowflake + AI vendor cert achievement for the AI Service team.
What comes next
The Gen AI specialty is a strong landing point. After completion, the AI Service team can explore:
- SnowPro Advanced: Data Engineer — if AI engineers want the full Snowflake advanced credential alongside the specialty
- Deeper Cortex project work — building production RAG systems and contributing to the Snowflake community as a Subject Matter Expert: learn.snowflake.com/en/pages/snowpro-sme-certification-program
Owner: Brylle Girang | Slack: #ai-learning
Last updated: 2026-04-01