GET /v1/summary reads 26 indicator series for a symbol and
timeframe, scores them against fixed thresholds, and returns a signed bias with a
confidence value, per-category directions, the conditions that fired and the raw
readings behind them — as one structured JSON object.
{ "success": true, "data": { "symbol": "EURUSD", "timeframe": "H1", "bias": "bullish", "bias_strength": "strong", "confidence": 0.8, "trend_score": 5.5, "momentum_score": 0.5, "volatility_score": 0.48, "signals": { "trend": "bullish", "momentum": "neutral", "volatility": "normal", "volume": "neutral" }, "key_levels": { "resistance": [1.0861], "support": [1.0819] }, "bullish_signals": ["Price above 200 SMA (long-term uptrend)", …], "bearish_signals": [], "neutral_signals": ["ADX 21.4 (weak/ranging market)", …], "volatility_info": […], "volume_info": […], "summary": "EURUSD on H1: Overall bias is STRONG BULLISH (confidence: 80%). Bullish factors: …", "recommendations": […], "key_values": { "bid": 1.08432, "ask": 1.08445, "rsi_14": 61.2, "macd_hist": 0.00021, … }, "updated_at": 1774620183 } }
Reading 26 series yourself is the easy half. The half that takes weeks is deciding what each reading means, agreeing a weighting, holding the two halves consistent as you add symbols, and keeping the verdict reproducible. This endpoint is that second half, pinned in one place.
26 indicator reads at 1× each, plus classification rules per indicator, a weighting table, an aggregation step, a narrative generator and a test suite that proves all of it still agrees with itself after the next change.
One GET with two parameters returns the bias, its strength, the
confidence, three scores, four category directions, the key levels, four signal
arrays, the narrative and all 13 raw readings that produced them.
key_values removes the follow-up callThese are the exact series the endpoint pulls from the indicator cache, grouped by the analyser that consumes them. Trend and momentum are the only two whose scores are added together to produce the bias; volatility and volume contribute observations and the per-category directions.
Moving-average structure, MACD state, ADX directional dominance and Parabolic SAR position. The price-versus-200-SMA check is the only condition in the whole endpoint worth ±2.
Oscillator bands. Each reading is classified against fixed numeric boundaries and contributes ±1 or ±0.5 — RSI above 70 and RSI below 30 are both scored, in opposite directions.
Band position and range. Produces the `volatility_info` strings and feeds `volatility_score`, but its score is not added to the bias total.
Flow confirmation. Produces `volume_info` and the per-category `signals.volume` direction; like volatility it does not move the bias total.
Why 26 and not 42. 42 is the number of individually addressable series on the platform. This endpoint reads the 26 listed above — the ones its four analysers have rules for. The two numbers are different measurements and are never presented as the same one.
Every condition adds a constant to its analyser's score. The trend and momentum totals are summed, and that one number selects the bias and its strength from the table below. The confidence value is a direct function of the same total — it is not a probability and it is not learned.
| trend_score + momentum_score | bias | bias_strength | confidence |
|---|---|---|---|
| total ≥ 4 | bullish | strong | min(0.9, 0.5 + |total| × 0.05) |
| 2 ≤ total < 4 | bullish | normal | min(0.75, 0.4 + |total| × 0.05) |
| −2 < total < 2 | neutral | normal | 0.3 + (1 − |total| ÷ 4) × 0.2 |
| −4 < total ≤ −2 | bearish | normal | min(0.75, 0.4 + |total| × 0.05) |
| total ≤ −4 | bearish | strong | min(0.9, 0.5 + |total| × 0.05) |
One condition is worth double. Price against the 200 SMA adds or subtracts 2. The golden/death-cross check, the short-term moving-average check, the aligned-EMA check, a MACD crossover and a directional ADX above 25 are each worth 1; a MACD histogram with no crossover and a mid-band RSI are worth 0.5. Parabolic SAR and a sub-25 ADX produce a signal sentence and no score at all.
One required parameter, one optional, one header. That is the whole contract — which is what makes it straightforward to wrap in an SDK, a scheduled job, an agent tool definition or a spreadsheet function.
symbol, the canonical name. timeframe, defaulting to H1. X-API-Key header. indicators permission scope. # the whole request curl -H "X-API-Key: YOUR_API_KEY" \ "https://tickatlas.com/v1/summary?symbol=EURUSD&timeframe=H1"
# read the structured fields, not the prose d = requests.get( "https://tickatlas.com/v1/summary", headers={"X-API-Key": KEY}, params={"symbol": "XAUUSD", "timeframe": "H4"}, ).json()["data"] print(d["bias"], d["bias_strength"], d["confidence"]) print(d["signals"]) # four category directions print(d["key_values"]["rsi_14"])
// bias transitions, not raw polling output const url = "https://tickatlas.com/v1/summary?symbol=EURUSD&timeframe=H1"; const headers = { "X-API-Key": KEY }; const { data } = await (await fetch(url, { headers })).json(); if (data.bias !== last.bias || data.bias_strength !== last.bias_strength) { notify(`${data.symbol} ${data.timeframe} → ${data.bias}`); }
The narrative is one field out of 19. Everything a program needs to act — the classification, the strength, the numeric scores, the per-category directions, the levels and the raw readings — is a typed field beside it, so nothing has to be parsed back out of the sentence.
data.
In the order the endpoint emits them. The envelope is
{ "success": true, "data": { … } }.
| Field | Type | What it carries |
|---|---|---|
| symbol | string | Canonical symbol. |
| timeframe | string | The timeframe analysed. H1 when the parameter is omitted. |
| bias | string | One of bullish, bearish, neutral. Lowercase — the uppercase form only appears inside the generated summary sentence. |
| bias_strength | string | strong or normal. strong is reached at a score of ±4; it is a separate field, not a prefix on bias. |
| confidence | number | Derived from the absolute score, rounded to 2 decimals. Capped at 0.9, and never below 0.3. |
| trend_score | number | Signed trend total, 1 decimal. One of the two inputs to the bias. |
| momentum_score | number | Signed momentum total, 1 decimal. The other input to the bias. |
| volatility_score | number | ATR_14 ÷ bid × 1000, clamped to 0–10 and rounded to 2 decimals. 0 when either value is missing. |
| signals | object | Per-category direction: trend, momentum, volatility, volume. Each is bullish / bearish / neutral at a ±0.5 boundary; volatility is always reported as normal. |
| key_levels | object | resistance and support, each an array. Populated from BB_upper and BB_lower — one value each, or an empty array when the band is unavailable. |
| bullish_signals | string[] | Every bullish condition that fired, across all four analysers, as readable sentences with the triggering number inlined. |
| bearish_signals | string[] | The same for bearish conditions. |
| neutral_signals | string[] | Conditions that fired but scored nothing, e.g. an ADX under 25. |
| volatility_info | string[] | Volatility observations. Informational only — these never move the bias. |
| volume_info | string[] | Volume observations, on the same informational footing. |
| summary | string | One generated paragraph: bias and confidence, then up to three bullish and three bearish factors, the leading volatility note and the leading entry of recommendations. Elided in the sample above. |
| recommendations | string[] | Generated strings. TickAtlas does not reproduce their wording in its own documentation — see the note below this table. |
| key_values | object | The 13 raw readings behind the verdict, so a caller never needs a follow-up indicator request. |
| updated_at | number | Epoch timestamp of the last indicator-cache refresh that fed this response. |
On recommendations and summary. The live endpoint builds
both, and summary appends the first entry of recommendations to
its closing sentence. TickAtlas supplies market data and computed analytics for
software use and does not give personalised investment advice, so this page documents the
two fields by name, type and origin and does not reprint their wording. Every sample value
shown elsewhere on this page is a descriptive reading or a classification.
| key_values | Source series |
|---|---|
| bid, ask | live quote cache |
| rsi_14, stochastic_k, mfi_14 | RSI_14, Stochastic_K, MFI_14 |
| macd_hist, adx, atr_14 | MACD_hist, ADX, ATR_14 |
| sma_20, sma_50, sma_200 | SMA_20, SMA_50, SMA_200 |
| bb_upper, bb_lower | BB_upper, BB_lower |
The same two-parameter request on M1, M5, M15, M30, H1, H4, D1. H1 is the default. A value outside this set is rejected with a 400 rather than quietly coerced, so a typo surfaces at the call site instead of in your output.
The same authentication, the same envelope and the same error shape as every other endpoint on the platform. Nothing about the aggregation changes how you call it.
Indicator values are served from Redis, so repeated reads inside one window return the same numbers without recomputing. Each timeframe is held just past the interval at which our data updates it.
cache_ttl: M1 180s → D1 2700sOne request header. The key must carry the indicators scope — the same scope the single-indicator endpoints require.
X-API-Key: tk_…Pass EURUSD; the canonical name is what comes back in symbol.
Each call counts once against your daily quota and costs five weighted units of pay-as-you-go credit. The weight is fixed and visible in your usage breakdown, so cost is predictable before you ship.
Weights below are the real usage multipliers, not estimates. Reach for the cheapest endpoint that answers your question, and keep the summary for the cases where the aggregate verdict is the thing you actually need.
The summary endpoint sits inside the developer API plans. Each call counts once against a daily quota; the 5× weight only sets its pay-as-you-go cost ($0.025 at the entry rate).
The response is shaped for programs: typed scalars for logic, arrays of readable sentences for display, and one prose field for the cases where a human is the consumer.
The whole data object is already a flat JSON record with typed fields and readable signal sentences, so it drops into a tool-call result with no reshaping step.
bias, bias_strength, confidence and the four signals directions map onto a badge, a meter and four category lights without any client-side maths.
Store bias plus bias_strength per symbol and timeframe, compare on the next poll, and act on the transition rather than on every response.
How the score is built, what the weights actually are, what the endpoint costs, and where the line sits between analytics and advice.
Four analysers run over 26 indicator series. Each condition that fires appends a sentence to a bullish, bearish or neutral list and adds a fixed amount to its analyser's score — ±2 for price against the 200 SMA, ±1 or ±0.5 for everything else. The trend and momentum scores are then added together, and that single total selects the bias and strength from a fixed ladder at ±2 and ±4. Volatility and volume conditions produce text and the per-category directions, but are not part of the total.
No. Every weight is a constant in the endpoint. There is no regime detector, and no indicator is promoted or demoted depending on whether the market is trending or ranging. The one asymmetry is that the price-versus-200-SMA condition is worth twice any other.
The summary endpoint is weighted 5× in the usage tier map, alongside the historical indicator endpoint. One call reads 26 series for the symbol and timeframe, runs all four analysers, derives the scores, the per-category directions, the key levels and the volatility score, and writes the narrative — server-side, in one round trip.
M1, M5, M15, M30, H1, H4, D1 — the same set every other indicator endpoint accepts. H1 is the default when the parameter is omitted, and anything outside the set is rejected rather than silently coerced.
Not for developer API keys. The summary entitlement check only applies to monitor/tool sessions, so an api-key request reaches the endpoint on any plan — including a pay-as-you-go account running on its $2.50 of starting credit. The key itself must carry the indicators permission scope.
A list of generated strings, derived from the trend and momentum scores, the volatility notes and the bias. TickAtlas publishes market data and computed analytics for software use; it does not provide personalised investment advice, and it does not reproduce the wording of that field in its own product documentation. The field is documented by name and type so your integration can handle it, and every other field on this page is descriptive.
Indicator values are served from a per-timeframe Redis cache, so two calls inside the same cache window return the same numbers and both cost 5× quota. Polling on the cadence of the timeframe you requested — once per new candle — is what changes the answer.
Start with one summary request on your $2.50 of starting credit, then reuse the same two-parameter call across every symbol and timeframe your product covers.