Junction 2026

Hackathon Tier 2published Preview page Onboarding PDF Live page

Event

Name

Type

Start

End

City

Venue

Website

Luma URL

Expected participants

Audience

Partner

Organizer name

Organizer email

Modal lead

Modal lead email

Hackathon specifics

Theme and tracks, who comes, what they build, tools and models, hardware needs, what they asked for, what Modal committed to. Drives the examples, the challenge and the page copy. Use "Draft specifics from context" to prefill from your emails, Luma, the website and Notion, then edit.

Offer

Program

Modal on site (talk, workshop, booth, mentors)

Credits per redemption ($)

Max uses (blank = expected redemptions x 1.2)

Code expiry

Credit code (create at modal.com/internal/credit-codes, paste here)

Prize

Workshop or talk

Support

Expected redemptions 540 (rate by audience), suggested max uses 648, exposure $200,000. Over $50k: needs the approvers in the Marketing credits guideline.
Code to create: amount $100, max uses 2000, expires 2026-11-16, note "junction-2026".

Challenge (blank = generated for Tier 2 and partner events) Pinned examples (slugs, comma separated; see catalog) Notes (internal)

Participants (0)

Luma: set the Luma URL above, then "Refresh stats" pulls registrations and check-ins (via the Luma API when LUMA_API_KEY is set, otherwise from Snowflake for events on Modal's Luma calendar). For organizer-run Luma pages without API access, export the guest list from Luma (Manage -> Guests -> Export CSV) and upload it here (columns: email, name, team).

Examples picked

Example	Why	Run
Run OpenAI-compatible LLM inference with Gemma and vLLM	Fits llm, code.	`modal run 06_gpu_and_ml/llm-serving/vllm_inference.py`
Build a coding agent with Modal Sandboxes and LangGraph	Fits agents, llm, agent, code, sandboxes.	`modal run 13_sandboxes/codelangchain/agent.py`
Run massively parallel evals on Modal Sandboxes	Fits agents, llm, agents, code, sandboxes.	`modal run 13_sandboxes/harbor_evals.py`
Create a Chatterbox TTS API on Modal	Fits llm, speech, about, voice.	`modal run 06_gpu_and_ml/text-to-audio/chatterbox_tts.py`
Run Flux fast on H100s with `torch.compile`	Fits vision.	`modal run 06_gpu_and_ml/stable_diffusion/flux.py`
Painting flowers with code	Fits llm, training, code, dojo.	`modal run 06_gpu_and_ml/paint_flowers/paint_flowers.py`
Deploy LangChain and LangGraph applications with LangServe	Fits agents, llm, agent, code.	`modal run 13_sandboxes/codelangchain/langserve.py`