Documentation Index
Fetch the complete documentation index at: https://scalarfield.io/docs/llms.txt
Use this file to discover all available pages before exploring further.
Overview
Scalar Field uses a unified credit system. Your subscription includes a monthly credit allowance. Credits are consumed by chat queries, agent execution time, agent LLM calls, data services, and other platform activity.
Your remaining credit balance is visible in the sidebar and updates in real time.
Subscription Tiers
| Plan | Monthly Price | Credits / Month | Key Includes |
|---|
| Free | $0 | 10 | Basic data access, limited queries |
| Pro | 80–175 | 440 – 970 | 200K context window, strategy agents, custom news agents |
| Ultra | 200–1,000 | 1,100 – 5,500 | 1M context window, higher agent limits, expanded data access |
| Enterprise | Custom | Custom | Unlimited queries, SSO, dedicated support |
Pro Plans
| Credits / Month | Monthly | Annual (per month) |
|---|
| 440 | $80 | ~$67 |
| 550 | $100 | ~$83 |
| 660 | $120 | ~$100 |
| 825 | $150 | ~$125 |
| 970 | $175 | ~$146 |
Ultra Plans
| Credits / Month | Monthly | Annual (per month) |
|---|
| 1,100 | $200 | ~$167 |
| 2,200 | $400 | ~$333 |
| 3,300 | $600 | ~$500 |
| 4,400 | $800 | ~$667 |
| 5,500 | $1,000 | ~$833 |
What Consumes Credits
| Activity | Credit Cost | Details |
|---|
| Chat queries | Varies by length | Based on conversation context size (tokens used) |
| Agent execution | 0.01 credits / second | Charged on total run duration of each execution |
Agent LLM calls (query_llm()) | Varies by model | tokens × 0.0001 × model_multiplier (see table below) |
| Email & alert delivery | 0.05 credits / email | Per notification sent by an agent |
| Data services (options analytics) | 0.01 credits / query | Per options analytics request |
Chat Query Credits
Each chat query consumes credits based on the size of the conversation context. Longer conversations with more messages use more tokens, which translates to more credits. The cost is computed as:
credits = tokens_used / tokens_per_credit
tokens_per_credit is set by your plan.
Agent Execution Credits
Every time a strategy agent runs, it is charged based on total execution duration:
credits = total_duration_seconds × 0.01
Agent LLM Calls (query_llm())
Python function: from scalarlib import query_llm, get_supported_models
query_llm() function in scalarlib. Credit cost depends on the number of tokens processed and the model selected:
credits = tokens × 0.0001 × model_multiplier
Supported Models
| Model (short alias) | Full Provider ID | Multiplier | Relative Cost |
|---|
gemini-2.5-flash | google/gemini-2.5-flash | 1× | Lowest (default) |
gpt-5.4-mini | openai/gpt-5.4-mini | 2× | |
grok-4.20 | x-ai/grok-4.20 | 3× | |
gemini-2.5-pro | google/gemini-2.5-pro | 4× | |
gpt-5.4 | openai/gpt-5.4 | 6× | |
claude-sonnet-4.6 | anthropic/claude-sonnet-4.6 | 6× | Highest |
model parameter:
from scalarlib import query_llm
# Default model (Gemini 2.5 Flash, 1x cost)
result = query_llm(
prompt="Which of these stocks are in the tech sector?",
response_schema={"tech_stocks": ["AAPL"]},
context=stock_list,
)
# Specify a model by short alias
result = query_llm(
prompt="Latest earnings surprises this quarter",
response_schema={"surprises": [{"ticker": "AAPL", "direction": "beat"}]},
web_search=True,
model="gpt-5.4", # 6x multiplier
)
get_supported_models() at runtime to see all available models and their provider IDs:
from scalarlib import get_supported_models
print(get_supported_models())
# {'gemini-2.5-flash': 'google/gemini-2.5-flash', 'gpt-5.4': 'openai/gpt-5.4', ...}
Credit Lifecycle
- Monthly reset: Credits reset at the start of each billing period. Unused credits do not roll over.
- Mid-cycle upgrades: If you upgrade your plan mid-cycle, your existing usage carries over to the new plan. You receive the higher credit limit immediately, but already-consumed credits are not refunded.
- Mid-cycle downgrades: Scheduled for the next billing cycle. Your current plan remains active until renewal.
- Exhaustion: When credits run out, chat and agent features are paused. You will see a banner prompting you to upgrade or wait for the next billing cycle. Trading strategies with active positions are not liquidated, but scheduled executions will not run.
Monitoring Usage
- Sidebar: Your credit balance and remaining credits are shown in the left sidebar, updated in real time via live streaming.
- Low balance warning: A warning banner appears in the chat input when credits are running low.
- Agent activation: If you have no credits remaining, you cannot activate new agents. Existing agents will pause execution until credits are available.
- Execution results: Each agent execution shows its
total_duration_seconds in the execution results panel, so you can estimate the credit cost of each run.
